Proceedings - Zurich, Switzerland 20-25 July 2003 ... - IARC Research

8th Infernational Conference on Permafrost
Zurich, Switzerland,
20 - 25 July 2003
Extended Abstracts
Reporting Current Research
and New Information
edited by
Wilfried Haeberli and Dagmar Brandova
Glaciology and Geomorphodynamics Group
Geography Department, University of Zurich
Switzerland

Preface
There is a well-established tradition to prepare and print the Proceedings of the International Conferences on Permafrost
(ICOP) in advance, enabling participants to take their volumes home from the meeting. As a consequence of this, papers
have to be submitted, reviewed, corrected and finally edited ahead of the conference and in time for all the work to be
published and delivered. In order to encourage presentation and discussion of ongoing activities and latest results, the
possibility was introduced for the 8th International Conference to present posters on current research and new
information. The present volume contains the extended abstracts of those posters which were personally presented at
ICOP 2003 in Zurich, Switzerland, 20 - 25 July 2003. Corresponding 2-page contributions had to be submitted in early
spring 2003 and were subject to a simplified yeslno-review process by the organizing committee. Our thanks go to the
conference sponsors, to the authors for their excellent collaboration, to Perscheng Assef and Margit Haeberli Vunder for
their help with editing and formatting, to Susan Braun, Marcia Phillips, Sarah Springman and Ivan Woodhatch for their
help with English smoothing, to Andi Kaab for organising the posters and to Martin Holzle for planning and final
preparation of the printing.
Wilfried Haeberli and Dagmar Brandova

Conference Sponsors
Principal Sponsor
International Permafrost Association (IPA)
Co-Sponsors
European Science Foundation (ESF)
International Commission on Snow and Ice (ICSI)
International Glaciological Society (IGS)
International Society for Soil Mechanics and
Geotechnical Engineering (ISSMGE)
Permafrost and Climate in Europe
in the 21st Century (PACE21)
Swiss Academy of Sciences (SANW)
Geoforum Switzerland of SANW
Glaciological Commission of SANW
Swiss Geomorphological Society of SANW
Swiss Academy of Engineering Sciences (SATW)
http://www,geodata.soton.ac,uk/ipa/
http://www.esf.org/
http://geowww.uibk.ac.at/research/icsi/
http://www.spri.cam.ac.uk/igs/home.htm
http://w,issmge.org/
http://www.cf,ac.uk/earth/pace/
http://www.naturalsciences.ch/
http://www.geoforum.ethz.ch/
http://www.sanw.glazko.sanwnet.ch
http://www.sanw,ch/exthp/sgmg/
http://w.satw.ch/indexd.html
Funding Sponsors
Main Funding Sponsor
Swiss Agency for Development and Cooperation (DEZA)
Co-funding Sponsors
Airbornescan, Switzerland
BHP Billiton Diamonds, Canada
Department of Geography, University of Zurich, Switzerland
GeobrugglFatzer Protection Systems, Switzerland
Helibernina, Switzerland
Institute for Geotechnical Engineering, ETH Zurich, Switzerland
Migros Culture-Percentage, Switzerland
Solexperts, Switzerland
Stump Bohr Drilling Company, Switzerland
Swiss Academy of Sciences
Swiss Academy of Engineering Sciences
Swiss Alpine Club
Swiss Cablecars
Swiss Federal Institute for Snow and Avalanche Research
Swiss Federal Office for Water and Geology
Swiss National Cooperative for the Disposal of Radioactive Waste

Extended Abstracts

Kamchatka: mountain permafrost, volcanoes and life
A.A.Abramov, S.N.Buldovich, *V.A.Shcherbakova, K.S.Laurinavichius, **E.M.Rivkina, L.Kholodov and D.A.Gilichinsky
Department of Geocryology, Moscow State University, Moscow, Russia
*Institute of Biochemistry & Physiology of Microorganisms, Russian Academy of Sciences, Pushchino, Moscow Region,
Russia
**Soil Cryology Laboratory, Institute of Physicochemical & Biological Problems in Soil Science, Russian Academy of Sciences,
Pushchino, Moscow Region, Russia
According to the NASA point of view, one way to have liquid
water on Mars at shallow depths would be through subglacial
volcanism. Such volcano-ice interactions are feasible
nowadays beneath the polar ice caps of Mars, or even within
the adjacent permafrost at the margins of the ice caps. This is
why one of the Earth’s models which may come close b
extraterrestrial environments is represented by active volcanoes
in permafrost areas. The main question is whether volcanoes
and associated environments can be ecological niche
for microbial communities, and whether these bacteria can
survive in permafrost. For this reason, our study was carried
out on the volcano Ploskii Tolbachik, situated in Kamchatka -
a unique peninsula in the east of Russia. It is one of the largest
modem volcanic regions with widespread glaciers and
permafrost at high altitudes. Large parts of the glaciers are
well studied, cold and situated at altitudes higher than 3000 m
a.s.1. (Kotlyakov 1997). Permafrost in the region, however,
has received little attention; on the basis of air temperatures,
continuous permafrost can be assumed to occur above 180CL
2000 m a.s.1. and to reach a thickness of 50400 m
(Zamolodchikova 1989). In a generalized way, and based on
new data obtained from drilling on the slopes of the volcano in
the summer of 2002, preliminary results relating to microbiological
and biogeochemical analyses are presented.
One 54 m deep borehole was drilled in 1978 at a site
where the Large Tolbachik Fissure Eruption (LTFE; Fedotov
1984) broke through in 1975176. In addition, we drilled a series
of boreholes and pit-holes at different altitudes between
800 and 2600 m a.s.1.onthe slopes of the active volcano
Ploskii Tolbachik. The boreholes were placed in different
landscape conditions, and had depths ranging from 5 to 15 m.
The permafrost samples for microbiological analysis were
obtained by slow rotary drilling without the use of solutions or
drilling mud. The surface of the extracted cores was trimmed
away with a sterile knife. Then the samples were immediately
plated on a specially-fitted nutrient medium. The main
part of the core was divided into sections, placed in presterilizedaluminumcans,sealedandplacedinfrozenstorage.
Samples were taken for ice content, gas, chemical and
textural composition. After completion of drilling, the boreholes
were closed during a week and then borehole temperatures
were measured.
Based on these field measurements and on other pub
lished data (Zamolodchikova 1989), it is possible to suggest
the following scheme of permafrost distribution on Ploskii Tdbachik
slopes (Fig. 1). The borehole 5/02, located in the forest
zone (-800 m asl.), has no permafrost. We can explain this
by the fact that the trees tend to retain snow. In fact, snow
depths in the forest can reach 2 m and more, whereas in the
absence of trees snow depth averages 0.5 to 0.8 m, and
hardly depends on wind redistribution. Enhanced snow depth
in forests effectively protects the soil surface from strong
winter cooling and enables seasonal freezing only. Borehole
6/02 drilled above the timberline at an altitude of 900 m a.s.1.
is in the zone of seasonal freezing, too. Because of lower
snow cover, however, the remains of seasonally frozen
ground with a thickness of 0.5 m were encountered at depth a
of 1.5 to 2 m.
Borehole “MARS-7/02” was drilled from the surface of a
tefra plateau at about I00 m a.s.1. in the belt of discontinuous
permafrost occurrence. The thickness of the active layer is
about I .5 to 2 m. The mean annual temperature of the frozen

ocks is between 0 and -1" and permafrost thickness up b
50-60 m. Borehole 30 drilled three years after LTFE at about
1000 m a.s.1. revealed 54 m of frozen rocks. It is interesting
to note that the textural description of this borehole indicated
combined piroclastic deposits from LTFE in the upper 16 in m
the summer of 1975. In 1978, these deposits had become
completely frozen (Andreev 1982). Model calculations show
that such deposits with a low heat conductivity need considerable
time for complete freezing (-8-9 yr). We cannot completely
explain this phenomenon yet.
Borehole "PIONER-1/02" was drilled at an elevation d
2600 m a.s.1. from the surface of a weathered basalt flow.
Permafrost here probably has a continuous distribution pattern
and a low mean annual temperature around -7 "C; its thickness
can reach 100-150 m. Local taliks can be found where
thermal waters exist and in the central part of the active volcano.
For better estimates of probable permafrost thickness in
the region, a numerical simulation using the program
"WARM" for the temperature field at the time of LFTE (1975
yr) was conducted. The model dimensions of the Ploskii Tolbachik
volcanic complex are 25 km long and 25 km deep.
The upper boundary condition is given by the mean annual
surface temperature calculated using an altitude lapse rate d
0.5"C/100 m. In glacierized terrain were taken the boundary
condition of Ill type (i.e. combination of mean annual air temperature
and coefficient of heat transfer between atmosphere
and rocks through ice with an estimated thickness of 100 m).
At the lower boundary, located on the depth 25000 m, was
set heat flux 80 mW/m2. On the depth 20000 m was set the
magma reservoir (3 km in diameter), with temperature
1200°C. Thermal properties of the rocks were taken from the
literature data (Komarov et al. 1996; Table I).
Table 1. Thermal properties of rocks used for simulation.
Basalt Volcanic dross
AT, W/m* K 2.9 0.3
hf. W/m*K 3.5 0.6
Cr, Kkal/m'*K 600 300
Cp, Kkal/m3*K 500 400
Q~, ua1lrn3 5000 1000
With the results of the simulation it is possible to conclude
that the volcanic eruption causes instantaneous temperature
rises of up to 1000-1200") but constitutes short-lived effects.
The main effects are concentrated at the surface and within
high atmospheric layers. Thus, it is not likely that such erup
tions will greatly reduce the thickness of the permafrost, any
effects will remain localized. Increased heat flux in areas d
modem volcanic activity, on the other hand, results in considerable
reduction of permafrost thickness underneath glaciers,
thereby probably inducing the formation of zones with
positive subglacial temperatures and basal glacier melting.
The subsurface temperature field at the the Ploskii Tolbachik
volcano has a non-stationary character, because it could not
fully adjust between eruptions.
Abramov, A.A. et al.: Kamchatka: mountain permafrost, volcanoes and ...
The existence of frozen rocks on the investigated volcano
is a function of elevation rather than latitude. As a consequence,
it should be attributed to mountain permafrost conditions.
In the investigated region, continuous permafrost with a
thickness of about 50 m starts to occur at an elevation higher
than 1400-1500 m a.s.1. Maximum permafrost thickness can
probably reach 200-250 m at 3000 a.s.1. m
The analysis shows thatfrozensamplesextractedfrom
borehole 7/02 and representing young volcanic deposits
contain viable microorganisms and, among them, therm@
philic anaerobic bacteria. Moreover, biogenic methane (up b
1100-1900 pICH4lkg soil) was found in these samples.
Thermophiles had never been found before in permafrost. The
present study therefore demonstrates that the only way for
thermophilic bacteria to appear within frozen volcanic horizon
is during eruption from volcanoes or from related subsurface
geological strata. The most important conclusion is that (1)
thermophilicbacteriamight survive in permafrost and even
produce biogenic gases, and that (2) terrestrial volcanic microbial
communities might serve as exobiological models for
hypothesesonexistingancientmicrobiocenoses,i.e., extraterrestrial
habitats that may possibly be found under anoxic
conditions around volcanoes on Mars or other planets.
ACKNOWLEDGEMENTS
This work was supported by the Russian Foundation of Basic
Research (grant 01-05-05-65043) and the NASA Astrobiology
Institute Program.
REFERENCES
Andreev, V. 1982. Permafrost in Tolbachik eruption region. Questions
of Kamchatka geography 8: 98-99 (in Russian).
Fedotov, S. (ed.) 1984. Large Tolbachik Fissure Eruption Kamchatka
1975-1976. Moscow: Nauka (in Russian).
Komarov, I. et al 1996. Heat- and mass-exchange properties of
rocks. In: E. Ershov (ed.), Lithogenous Geocryology. Moscow
State University publisher (in Russian).
Kotlyakov, V. (ed.) 1997. World atlas of snow and ice resources .
Moscow: Nauka (in Russian).
Zamolodchikova, S. 1989. Kamchatka region. In: Ershov, E. (ed.),
Geocryology of USSR. East Siberian and far East: 380-389.
Moscow: Nedra (in Russian).
2

8th international Conference on Permafrost Extended Abstracts an Current Research and Newly Available Information
Verifying modelling approaches: high mountain permafrost and its
environment
M. Avian
Department of Geography and Regional Sciences, University of Graz, Austria
INTRODUCTION
The Ankogel Mountains are a highly glacierized area in the
easternmost part of the Hohe Tauern Range in Carinthia
(Fig. I). The test site encloses 98 km* including the main
peaks of the Ankogel (3246 m) and the Hochalmspitze
(3360 m). Glaciers cover an area of 12.1 km2 and shows
wide retreat zones.
Remote sensing data are the basis for the mapping of
the distribution of snow patches, vegetation and rock glaciers.
Snow patches were mapped by using colour orthophotographs
(as at: August 1998), Keller (1994) shows an
investigation, in which a significant correlation between
493 snow patches and the occurrence of permafrost was
detected. Infrared orthophotographs are a common data
set for investigating vegetation. A useful classification in
connection with the investigation of permafrost could only
be carried out by visual interpretation. In connection with
their percentage of soil cover, three classes of vegetation
were used: uniform vegetation (over go%,), transition
vegetation (20 to go%,), and island vegetation (below
20%). The background is the assumption that a solid cover
of vegetation excludes permafrost in the underground. The
occurrence of active rock glaciers is an obvious sign for
permafrost.
RESULTS
Figure 1. Location of the study area in Austria.
METHODOLOGY
The distribution of permafrost was modelled by an adaptation
of the programme PERMAKART, using the GIS Arc-
Info (Haeberli et al. 1996). The main point is to receive information
about slope, aspect and elevation. The result is a
supposed distribution of permafrost in the study area with
the classes “permafrost probable”, “permafrost possible”
and “no permafrost”. The modelling of potential incoming
short wave radiation (PSWR) was done by using the programme
ERDAS-Imagine. Shadows that are cast by topographic
features on the surrounding surface were not calculated.
As the distribution of exposition is homogenous it is a good
basis for further analysis. The modelled classes of the distribution
of permafrost were renamed for the reason of empirical
results in the Hohe Tauern Range (Lieb 1998) from
“permafrost possible” to “potential sporadic permafrost
(PSP)” and from “permafrost probable” to “potential discontinous
permafrost (PDP)” (see Tab. 1). The model of
the PSWR shows a satisfying overall view and a good correspondence
of areas with a PSWR under 53 kJ/m2a and
PDP.
The main part of the study is to compare the results of
the modelling with the information received from the study
area. 684 snow patches with an area of altogether 108 ha
were counted in the study area (average altitude: 2635m).
At first sight there is no significant relation between PDP
and the occurrence of snow patches (Tab. 2), but a closer
look shows that nearly all snow patches which are not situated
in potential permafrost areas, occur in direct neighborhood
with PDP and PSP.
3

29.3% of the study area are covered with vegetation,
the relation between potential permafrost and vegetation
provides a clear result: although there is a high percentage
of soil cover with vegetation in this high mountain area,
only a marginal zone of overlapping with PDP in an area of
0.9 hectares and 0.06% can be seen (Table 3). Rock glaciers
can only be found in the southern part of the study
area. Seven are active, one is inactive and eight are fossile.
The model shows a good correlation with the catchment
area of blocks, but does not mainly correspond with
the lobes of the rock glaciers.
Table 1. Areas of generated permafrost classes.
area in km2 area in %
no permafrost
PDP 24.0 24.6
sum 97.6 100.0
Table 2. Snow patches according to potential permafrost areas.
Avian, M.: Verifying m ~ approaches: ~ high ~ mountain ~ permafrost.. l ~ ~ ~
1 area in km2 I area in %
no nermafrnst I 0.37 "" I 84.5 - ."
PSD 0.37 34.0
PDP 0.34 31.5
sum I 1.08 I 100.0
Table 3. Areas of vegetation classes and overlapping with potential
permafrost areas.
area in area in Overlapping %
km* % PDP PSP
uniform veg. 17.9 62.3 0.05 1.33
transition veg 8.6 0.12 30.1 9.86
island veg. 2.1 2.30 7.6 19.06
sum 28.6 100.0
The final step is to generate a most probable lower borderline
for permafrost including all results with a last visual
interpretation of the geological situation (coarse debris).
The adaptation uses the modelling of potential permafrost
areas as a basis, the results of the distribution of vegetation
correct the limit as well as the distribution of snow
patches. A visual interpretation of the orthophotographs includes
areas of coarse debris in the neighborhood of potential
permafrost areas. The modelling of PSWR is only a
check in a higher scale. The final result is an area of 35.16
km2, that are 36% of the study area, which can be seen as
potential permafrost areas. In April 2002 BTSmeasurements
were carried out in two small test sites in
the south east of the study area. The results correspond
very well with the most likely lower borderline for permafrost.
I. Although there is a superficial discrepancy between
the assumption and mapping results of
snow patches, a closer look provides valuable improvement.
2. Closed vegetation cover excludes potential permafrost
in the study area with a probability of
99,96%.
3. All catchment areas of active rock glaciers are
situated in potential permafrost areas.
4. The result of modelling the potential incoming
short wave radiation shows a good connection to
potential permafrost.
Crosschecking of results shows good correlations between
the modelling results and the results from mapping
for the Eastern part of the Alps. Valuable improvements
can be derived in using the distribution of snow patches
and geological patterns.
ACKNOWLEDGEMENTS
I wish to thank Mr. Gerhard Lieb and Ms. Michaela Nutz for
critical review, also Mrs. Christine Lazar for improving the
English and Mr. Stefan Lieb, Mr. Markus Frei, Mr. Wolfgang
Sulzer as well as Mr. Alexander Avian for critical
comments.
REFERENCES
Haeberli, W., Hoelzle M., Dousse, J.-P., Ehrler, C., Gardaz, J.-M., Imhof,
M., Keller, F., Kunz, p., Lugon, R. and Reynard, M. 1996.
Simulation der Permafrostverbreitung in den Alpen mit geographischen
lnformationssystemen. Arbeitsbericht NFP 31,
Hochschulverlags-AG an der ETH Zurich ,58 p.
Keller, F. 1992. Automated mapping of mountain permafrost using the
programm PERMAKART within the Geographical Information
System ARC-Info. Permafrost and Periglacial Processes, Vol. 3(2):
133-138.
Keller, F. 1994. lnteraktionen zwischen Schnee und Permafrost, Eine
Grundlagenstudie im Oberengadin. Versuchsanstalt fur Wasserbau,
Hydrologie und Glaziologie der ETH Zurich, 145 p,
Lieb, G.K. 1998. High-mountain permafrost in the Austrian Alps. -
Permafrost. 7th International Conference on Permafrost. Yellow-
CONCLUSION
The model of the distribution of permafrost provides wide
areas of potential permafrost, including almost the entire
glacier areas. All the relations can be summarised as follows:

The role of cosmoplanetary climate cycles in Earth’s cryolithosphere
evolution
V. T. Balobaev and V. V. Shepelev
Melnjkov Permafrost Institute, $0 RAS, Yakutsk, Russia
We believe that different climate-forming global factors
have a resonance type of relationship. The conceptual
model of manifestation of this effect on the global climate
system is shown in Figure 1. This model is based on the
assumption that periods and amplitudes of the temperature
variations caused by separate global factors are constant
in time. The presented model takes into account four
main climate-forming factors: one cosmoplanetary variation
with a period of 200 Ma and three astroplanetary
variation with periods of 100, 40.7, and 20 ka. The variation
curve of the longer cycle is taken as the base level for
variations of the shorter climate-forming cycle. The superposition
of harmonics involved in the calculation of
climate-forming cycles highlights distinct peaks corresponding
to periods of phase coincidence of individual temperature
variations. This effect, which we called thermoresonance,
causes significant periodic cooling and warming
of the global climate.
(k
k.
E 4
lntcrval helwecn extrcrnc wanning pcnuds. ka
40 **
80 ,
I
Figure 1. Conceptual model of the integral dynamic correlation
between the main global factors of climate formation during the past
150 ka. (1) curve of a 20-ka cycle caused by precession of the Earth’s
axis; (2) curve of a 40.7-ka cycle caused by variations in the obliquity
of the rotation axis with respect to the ecliptic plane; (3) curve of a
100-ka cycle caused by variations in the eccentricity of the Earth’s
orbit: (4) curve of an about 200-Ma cycle related to the rotation of the
solar system around the galactic centre; (5) area of optimal climatic
conditions; (6) periods of extreme warming; (7) periods of extreme
cooling.
5
The amplitudes of harmonics strongly vary and the
highest of them, especially those of 40-ka and 20-ka
cycles, distinctly indicate the resonance amplification.
Since the Earth’s climate system is excited and less stable
during the strongest temperature decrease or increase in
the 100-ka cycle, the 40- and 20-ka cycles exert a
resonance effect. It should be noted that extreme cold
periods in our model generally agree with glacial periods
in the Late Pleistocene reconstructed from seafloor
sediment data (Steig 1998).
Using the proposed concept of the thermoresonance
effect, an attempt can be made to reconstruct the climate
for the whole Phanerozoic Eon spanning the past 570 Ma.
Figure 2 shows a scheme constructed with consideration
for the thermoresonance effect between two main
cosmoplanetary climate-forming factors with periods of 1.3
Ga and 200 Ma and spanning the past 700 Ma of the
Earth’s evolution.
Of special interest for permafrost-researchers is to
reveal the possibility of formation and dynamics of
cryogenic processes in different periods of the
Phanerozoic which was relatively warm in the Earth’s
history that had served as the base for a vigorous
development of a biosphere. The preceded Proterozoic
was significantly colder, what determined the existence of
large glaciations whose traces are found in sediments of
this epoch. The Cambrian (570-500 Ma ago), that is the
first period of the Phanerozoic, was warm and without
permafrost what is connected not only with 200-Ma cycle
(Fig. 2) but with favourable location of lithospheric plates
on the Earth’s surface as well. The largest continent of
that time, Gondwana, was located in subtropical latitudes,
and smaller plates were in the equator zone. The poles
were located in water areas of open oceans, and negative
temperatures were not observed on the Earth’s surface of
the whole planet.
In the Ordovician (500-440 Ma ago) the planetary
climate was significantly colder (Fig. 2). By the end of this
period, Gondwana transferred to the South Pole. This
determined the initiation of a thick continental glaciation,

Balobajev, V.T. et al.: The role of cosrnoplanetary climate G~GIES..
whose traces occur in Africa. Frozen grounds developed
at a third part of Gondwana (Africa and South America).
Figure 2. Scheme of the correlation between two main cosmoplanetary
factors responsible for the Earth's global climate. (1) curve
of the 1.3-Ga cycle caused by rotation of our galaxy around the megagalactic
centre: (2) curve of a 200-Ma cycle caused by rotation of the
solar system around the galactic centre; (3) area of optimal climatic
conditions: (4) extreme warming period; (5) extreme cooling periods.
The water areas of the North Pole could have seasonal
icings only. The global warming occurred in the Silurian
(440-410 Ma ago). The Gondwana continent left the South
Pole and transferred to the equator. Thin frozen grounds
could occur just on a small part of this continent (South
America). The area of their development and the thickness
increased by the end of the Silurian due to drifting of
Gondwana to the South Pole.
The cycle of global warming (440-300 Ma ago)
coinciding with the periods of the Devonian and the
Carboniferous was one of the warmest cycles on the
Earth. It was accompanied by a vigorous development of
vegetation. The addition of small lithospheric plates to the
second supercontinent, Laurasia, occurred in this epoch,
and at the end of this epoch Gondwana joined with
Laurasia into a single supercontinent Pangea. There was
no glaciation though, Gondwana and later a significant
part of Pangea were located near the South Pole
throughout this period. However, considerable daily and
annual fluctuations of temperature and its transitions
through 0°C was to be observed due to a large area of the
continent and location of its parts in high latitudes. At the
end of the Carboniferous (350-285 Ma ago), when the
global temperature began to decrease, negative mean
annual temperatures and frozen grounds appeared in the
south part of Pangea. Then the continental glaciation
began near the South Pole (Antarctica, South Africa,
Australia). The strongest cooling of the Earth
wasobsewed at the boundary of the Permian (285-230 Ma
ago) and the Triassic (230-195 Ma ago). It was connected
with the global climate cooling and unfavourable location
of Pangea relative to the poles. The south of Pangea was
covered with ice (Antarctica), and the south of the
continent contained permafrost.
The subsequent cycle of a global warming occurred in
the Jurassic (195-137 Ma ago) and the Cretaceous (137-
67 Ma ago, Fig. 2). By that time Pangea was divided into
Gondwana and Laurasia and then into smaller blocks, of
which present continents were formed later. The reasons
that Gondwana moved to the latitudes of 20" from the
South Pole and that both poles of the Earth occurred in
the ocean resulted in a greater increase of the global
temperature. Mean global temperature at that time was
8°C higher than the present. All of this eliminated the
possibility of formation of permafrost on the Earth and
glaciation centres.
At the end of the Cretaceous, a new cold epoch developed
on the Earth. Compared to the Late Cretaceous, temperature
decreased by 2-4°C at the middle of the Paleogene
(65-23 ka) and was 44°C higher than the present at
the end of the Paleogene. Fast decrease of the global
temperature was due to the fact that Antarctica firmly
established at the South Pole. Formation of the Antarctic
circumpolar stream at the beginning of the Neogene
resulted in glaciation of Antarctica, which continues to the
present (Monin and Shishkov 1979). In the northern
hemisphere first sheet glaciations are dating by 3 Ma ago,
and the ice cover in the Arctic Ocean appeared 0.7-0.8 Ma
ago. Probably, at the same time the permafrost formed in
the northern hemisphere.
Therefore, positive phase of the 1.3-Ga climate cycle,
including practically the entire Phanerozoic, is terminating.
This indicates that the next 500-600 Ma on our planet will
be a kingdom of cold and ice. However, this does not
exclude short-term warming periods related to astro- and
geoplanetary temperature-forming factors and their resonance
amplification.
REFERENCES
Monin, A.S. and Shishkov, Yu.A. 1979. Climate history (Istoriya
klimata) (in Russian). Leningrad: Gidrometeoizdat.
Steig, E. 1998. Synchronous climate changes in Antarctica and the
North Atlantic. In Science 282: 92-95.
6

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Terrestrial biosphere modelling in permafrost regions
C. Beer, *W. Lucht, **C. Schmullius and M. Gude
Department of Geography, University of Jena, Germany
*fofsdam Institute for Climate lmpact Research, Potsdam, Germany
**Department of Geography, University of Jena, Germany
INTRODUCTION
Besides C02 inversion modelling, Dynamic Global Vegetation
Models (DGVMs) are suited for investigating the terrestrial
component of the global carbon balance. These
models are icreasingly of particular interest due to their
potential for predicting the impacts of the biosphere upon
future concentrations of atmospheric C02. Because water
availability in plants is one of the key stimuli for vegetation
growth, DGVMs combine the carbon with the water cycle.
Our research is conducted within the framework of the
SIBERIA-II project funded by the European Commission
(Contract No. EVG2-2001-00008, http://www.siberia2.unijena.de).
In this work, the impact of hydrological conditions
on the carbon cycle in permafrost regions of Northern
Eurasia is investigated. For this purpose, a permafrost
module built upon previous work of Venevsky (2001) following
the equations of Anisimov et al. (1997) is presented
as an extension of the Dynamic Global Vegetation Model
LPJ (Lund-Potsdam-Jena; Sitch 2000).
MODEL DESCRIPTION
LPJ uses 4 carbon and 2 soil layers (0.5 m and IS m
depth). Input of LPJ are basic climate and soil data (air
temperature, precipitation, sunshine duration, soil texture)
on a spatial resolution of 0.5" times 0.5". To provide dynamic
variables of the carbon cycle such as Net Primary
Production (NPP) or Leaf Area Index (LAI) in a daily time
step, LPJ begins by computing hydrological variables such
as soil moisture and evapotranspiration.
In permafrost regions the hydrology is completely altered
by our new permafrost module. Soil moisture is
strongly dependent on daily thaw depth in the present. As
a simplification, we assume that the course of daily thaw
depth and air temperature resembles a sine curve. This
leads to the following equation for calculating the daily
thaw depth:
Zthaw - daily thaw depth
L a x - max thaw depth
t - current day
tspring - start of spring
zpos - duration of days while T > 0
T - air temperature
The zonation of permafrost (Fig. 1) is obtained by assuming
that TTOP (Temperature at TOP Of the Permafrost)
is below 0 "C in permafrost regions following Smith
(2002). This zonation will be compared with that obtained
by means of the frost-index of Anisimov and Nelson
(1 997).
OUTLOOK
To investigate the impact of permafrost hydrology on the
terrestrial carbon cycle, different outputs of LPJ with and
without the permafrost module are compared with in situ
measurements or Earth Observation data, e. g., growing
stock volume, NPP, or LA1 in Siberia between 90" and
I IO"East. A lower value of biomass and NPP is expected
when permafrost hydrology is taken into consideration,
and it is anticipated that more reliable values for NPP and
NEP (Net Ecosystem Production), can be obtained, which
in turn will lead to a more precise global carbon balance.
7

Beer, 6. et al,: Terrestrial biosphere modelling in ~ ~r~~~fr~st
regions..
REFERENCES
Anisimov, O.A. Shiklomanov, N.I., and Nelson, F.E. 1997. Global
warming and active-layer thickness: results from transient general
circulation models. Global and Planetary Change 15: 61-77.
Anisimov, O.A. and Nelson, F.E. 1997. Permafrost zonation and climate
change in the northern hemisphere: results from the transient
general circulation models, Climatic Change 35: 241-258.
Sitch, S. 2000. The role of vegetation dynamics in the control of atmospheric
COe content. Dissertation, Lund University, Lund,
Sweden
Smith, M. W. and Riseborough, D.W. 2002. Climate and the limits of
permafrost: A Zonal Analysis. Permafrost and periglacial Processes
13: 1-15.
Venevsky, S. 2001. Broad-scale vegetation dynamics in north-eastern
Eurasia - observations and simulations, PhD Thesis, Universitat
fur Bodenkultur. Wien.

Combining water chemistry and geophysical investigations with assessment
of chemical denudation rates in the Latnjavagge drainage basin, arctic-oceanic
Swedish Lapland
A. A. Beylich, E. Kolstrup, *U. Molau, **T. Thyrsfed, * W. Linde, L. B. Pedersen and ****D. Gintz
Department of Earth Sciences, Environment and landscape Dynamics, Uppsala University, Uppsala, Sweden
*Botanical Institute, Plant Ecology, Gothenburg University, Gothenburg, Sweden
**Harbacken, Stavby, Alunda, Sweden
***Department of Earth Sciences, Geophysics, Uppsala University, Uppsala, Sweden
““*Institute for Geological Sciences, AB Hydrogeology, Free University of Berlin, Berlin, Germany
In I999 a monitoring programme was started in an arcticoceanic
drainage basin, Latnjavagge (9 km2, 950 - 1440
m a.s.1.; 68”20’N, 18”30‘E), with the objective to quantify
the sediment budget and the relevant denudation processes
as dependant on environmental parameters
(Beylich in press). With regard to chemical denudation
rates several factors such as topography, climate, lithology,
regolith thickness and ground frost are of importance,
so that as a consequence, different methods need to be
combined in order to assess the chemical denudation.
Once an integrated approach is applied, it is possible to
analyze chemical denudation rates for clearly-defined
landscape units such as confined fluvial drainage basins
(Beylich 2002).
The Latnjavagge drainage basin in the periglacial
Abisko mountain area is characterized by a homogeneous
lithology (mica shists). Water balance, water chemistry
and radio magnetotelluric geophysical investigations were
integrated with assessment of regolith thickness along
selected profiles as well as of ground frost conditions
within the catchment. In combination with direct field observations,
the geophysical profiles demonstrated the
presence of very thin regolith in most of the investigated
areas, yet in some parts the bedrock was located deeper
and was not detected locally at 40 m depth. The low resistivities
found along the profiles in the geophysical investigation
showed that frozen ground was not found around
1000 m a.s.1. in late August (Beylich et a/. subm.b). Permafrost
in the area seems to be restricted to areas above
1000 m a.s.1. Water chemistry and total dissolved solids
(TDS) of precipitation, snowpack and surface water have
been analyzed on samples collected at 25 selected sites
within Latnjavagge (Beylich et a/. subm-a). Additionally,
daily discharge and yield of dissolved solids have been
calculated for several subcatchments (Beylich et a/.
submx). Water laboratory analyses of 205 samples’taken
in the field included the components Ca2*, Mgzt, Na+, Kt,
Fe2+, Mn2+, CI-, Nos-, so42- and Po$, with S042- and Can+
being the dominant ones in the surface water (Beylich et
a/. subm.a). The salt concentrations and chemical denudation
in this arctic-oceanic periglacial environment were
found to be low with a mean annual chemical denudation
rate for the entire catchment of 5.4 ff(km2 yr) (Beylich et a/.
subm.c).
Comparison of the water chemistry data from different
sampling sites and subcatchments by means of cluster
analysis shows that there are significant differences within
the area: In areas of shade, longer snow cover, frozen
ground and thin regolith, the salt concentrations over the
summer were perceptible but so low that salts brought into
the basin from precipitation and melting snow could be
clearly detected in the surface water. The highest salt
concentrations were found at a radiation-exposed west
facing, vegetated, moderately steep slope with relatively
thick regolith that was thawed already from the time of
snowmelt in early June. In such well-drained sites with
continuous subsurface water flow a maximum of contact
between water and mineral particles can take place
(Beylich et a/. subm.a). The concentration values revealed
differences in the rate of thawing of frozen ground between
shaded areas andlor areas at higher altitude on the
one hand, and radiation-exposed areas on the other. A
comparison with the results from Karkevagge a few kilometres
to the northwest as well as from other periglacial
locations indicates that the chemical denudation values
from Latnjavagge are fairly representative of periglacial
oceanic environments, more so than the values from the
Karkevagge catchment (Beylich et a/. subm.a). The investigation
in Latnjavagge stresses the importance of spatial
variability within even small catchments of homogeneous
lithology as it demonstrates that salt concentrations from
different subareas can differ substantially depending on
exposure to radiation, duration of snow cover and frozen
ground conditions, regolith thickness, and also on vegetation
cover and slope angle as factors steering water turbulence
and retention of drainage (Beylich et a/. subm-a).
9
REFERENCES
Beylich, A. A,, 2001. Slope denudation, streamwork and relief development
in two periglacial environments in East Iceland and
Swedish Lapland. Transactions, Japanese Geomorphological
Union, 22 (4), C-23

Beylich, A. A. et al.: Combining water chemistry and ~ ~ ~ investigations ~ ~ ~ s...
i c ~ l
Beylich, A.A. in press. Sediment budgets and relief development in
present periglacial environments - a morphosystem analytical approach,
Hallesches Jahrbuch fur Geowissenschaften, A.
Beylich, A.A., Kolstrup, E., Thyrsted, T., Gintz, D., subm.a. Water
chemistry and its diversity in relation to local factors in the
Latnjavagge drainage basin, arctic oceanic Swedish Lapland,
Geomorphology.
Beylich, A.A., Kolstrup, E., Thyrsted, T., Linde, N., Pedersen, L.B.,
Dynesius, L., subrn-b. Chemical denudation in arctic-alpine
Latnjavagge (Swedish Lapland) in relation to regolith thickness as
assessed by radio magnetotelluric (RMT)-geophysical profiles,
Geomorphology.
Beylich, A.A., Molau, U., Luthbom, K., Gintz, D., subm.c. Rates of
chemical and mechanical fluvial denudation in an arctic-oceanic
periglacial environment, Latnjavagge drainage basin, northern-

8th International Conference on Permafrost Extended Abstracts on Current. Research and Newly Available i n f ~ ~ ~ ~ ~
Analysis of permafrost thickness in Antarctica using VES and MTS data
E. Borzotta and *D. Trornbotto
Unidad de Geofisica, *Unidad de Geocriologia
lnstitufo Argentino de Nivologia, Glaciologia y Ciencias Ambientales (IANIGLA)
CONICET, Casilla de Correo 330, 5500 Mendoza, Argentina
Frozen ground thicknesses measured in the NE of the
Antarctic Peninsula (Seymour and James Ross Islands;
Fig. I) and estimations of permafrost thicknesses for the
same areas are being compared and analyzed using (a)
vertical electrical soundings data (VES), (b) mean annual
air temperatureland (c) steady geothermal heat flows as
inferred from magnetotelluric soundings (MTS).
In the present study, as part of a more exhaustive
study in progress, estimations of steady heat flows are
made, which could contribute to a better understanding of
the cryogenic processes in the area.
Figure 1. Antarctic Peninsula. The frames indicate the zones where
permafrost studies were performed (Seymour and James Ross Islands).
Seymour and James Ross Islands are located in the
Weddell Sea, next to the Antarctic Peninsula, at about
64"s and 57"W. Both are part of the Larsen Basin located
eastward of the Antarctic Peninsula. Upper Cretaceous
marine sediments outcrop in the south portion of Seymour
Island and Tertiary sediments at the north. This island is
free of ice cover. James Ross Island is a Plio-Pleistocene
volcanic island with recent activity at its eastern border
and around 70% of its surface is covered by glaciers. At
Seymour Island, a mean annual air temperature (MAAT)
of -9.4"C was estimated at 200 m a.s.1. (upper terrace)
from a ten-year temperature recording (1971-1980). Temperatures
at 10 m depth were measured during the 1976-
1981 period in the snow on James Ross Island (Dalinger
rl"S
Dome and Mt. Haddington), which let to estimate a MAAT
of -13°C at about 1410 m, on average, and -5°C at 35 m
a.s.1. where geophysical studies were conducted later by
Fukuda and others (Fukuda et al. 1992).
-The first magnetotelluric soundings (MTS) in Seymour
Island were carried out in its northern section (upper terrace)
in 1979 and 1980 (Fournier et al. 1980, Del Valle et
al. 1982), describing the Larsen Basin only below 600 m
depth; thus it was not possible to study the permafrost.
Results showed a very conductive layer (saline) below the
frozen ground with 40 Siemens of conductance. A basin
basement at 5.2 km depth is also shown by these data
and a conductive layer (possible asthenosphere) is suggested
in the upper mantle at longer periods with a top at
63 km depth.
In 1992, three MT soundings were made in James
Ross Island (Brandy Bay and Hidden Lake). Results
showed a conductance below the frozen ground one order
of magnitude higher than below the frozen ground at
Seymour Island: 40 Siemens in Seymour Island, 315 Siemens
in James Ross Island. As in the case of Seymour
Island, a possible asthenosphere is also suggested for
James Ross Island with a top at 61 km depth. Distortions
present in the apparent resistivity curves and the volcanic
nature of this island, with recent activity, lead us to suspect
the existence of a possible magma chamber seated
between 6.8 km and 14 km depth in the crust at NW of
James Ross Island. These electromagnetic studies performed
in Seymour and James Ross Islands were not able
to describe the permafrost directly.
Vertical electrical soundings (VES), particularly dedicated
to studying the permafrost, were conducted on
Seymour and James Ross Islands in the summer 1989-
1990 (Fukuda et al. 1992). These studies were carried out
at different terraces observed by these authors. In
Seymour Island, 200 m of frozen ground thickness was
estimated in the upper terrace (around 200 m a.s.l.), 105m
in the middle terrace (around 50 m a.s.1.) and 35m in lower
terrace at 5 m a.s.1. Maximum resistivities of about 720 to
800 8m were measured in frozen ground by these
authors (Fukuda et al. 1992). In a general context, the
thickness of the active layer in Seymour Island could be
11

estimated at 0.35 to 0.60m. In the same campaign, four
VES were carried out in the ice-free northwestern region
of James Ross Island. Shallower thicknesses of frozen
ground than those estimated at Seymour Island were estimated
in this case: 40 to 45 m in the upper terraces
(around 21 to 35 m a.s.l.), 3.4 m in the middle terrace
(around IOto 17 m a.s.1.) and 5.8 m in the lower terrace at
about 3 to 5 m a.s.1. Resistivities between 1000 and 4000
m were obtained, with a thickness of the active layer
approximately 0.70 to 1.45 m (Fukuda et al. 1992).
From the depths of conductive layers in the crust and
the upper mantle determined by MT soundings, a gross
estimation of the regional geothermal heat flow - in steady
condition - may be obtained, and consequently, the geothermal
gradient and the permafrost thickness in equilibrium
to the present MAAT can be estimated. In the present
current study at Seymour and James Ross Islands,
the frozen ground thickness - measured using VES - and
the estimated permafrost thickness, obtained considering
homogeneous medium and steady heat flows inferred
from MT soundings results, are being compared and analyzed.
Empirical exponential formulas correlate the geothermal
heat flows and the depths of conductive layers determined
using MTS (Adam 1978). These expressions are
based principally on information published in Adam
(1976), corresponding to the East of Europe and Russia,
involving platforms and geo-dynamically active regions.
Considering 63 km as the suggested depth of the asthenosphere
in Seymour Island (also suggested by the MTS
performed in James Ross Island), a heat flow of 78
mWlm2 is estimated in this island (lpcallcm2sec. 42
mWlm2). If the magma chamber, suggested by the MT
study, really exists in James Ross Island its presence will
increase the heat flow at surface. In fact, considering the
top of this chamber at 6.8 km depth - according to MT results
- a heat flow of 148 mWlm2 is estimated using the
expressions mentioned. The heat flow values, thus estimated,
should only be considered as approximations of
the steady geothermal heat flows corresponding to these
regions. It is important to mention the presence of a rift at
the Bransfield Strait located 200 km to northwest from
Seymour Island, where heat flows higher than 220
mWlm2 were measured.
In order to estimate the geothermal gradient in the
permafrost of Seymour and James Ross Islands, a thermal
conductivity obtained in the laboratory (I.883 Wlm OK)
was used corresponding to two samples of frozen sands
with about 87-94% of water. In this way a geothermal gradient
of 0.04 "Clm is estimated for Seymour Island and
0.079 "Clm for James Ross Island. Taking into account
the MAAT for both islands: -9.4 "C for Seymour Island at
200 m a.s.l., (where the VES study gave 200 m of frozen
ground thickness) and -5.0 "C for James Ross Island at
about 35 m a.s.1. (where the VES study gave 40 to 45 m of
frozen ground thickness), we obtain 235 m and 63 m as
the estimated permafrost thickness corresponding to
Seymour and James Ross Islands respectively.
Borrotta, E. et.al.: Permafrost thickness in Antarctica usirrg.,
Because of the high conductivity below the frozen
ground put in evidence at both islands by MTS, the presence
of a cryopeg at the base of the frozen ground is very
probable. Therefore, the estimations of the permafrost
thickness made in these islands (235 and 63 m) are consistent
with the thickness of frozen grounds obtained by
VES (200 and 45 m). The differences could be attributed
to a cryopeg.
Although the permafrost thickness in equilibrium, thus
estimated, may contain uncertainties, its comparison with
the frozen ground thickness - corresponding to both
studied islands - seems to suggest a state approaching
equilibrium. This result means that the soil would have
been uncovered by ice and with approximately stable climatic
and geological conditions for a period of time long
enough to reach this state.
REFERENCES
Adam, A. 1976. (ed.). Geoelectric and Geothermal Studies (east -
central Europe, Soviet - Asia). KAPG Geophysical Monograph.
Akademiai Kiado, Budapest.
Adam, A. 1978. Geothermal effects in the formation of electrically
conducting zones and temperature distribution in the Earth. Phys.
Earth Planet. Inter. 17: 21-28.
Del Valle, R.A., Demichelli, J., Febrer, J.M., Fournier, H.G., Gasco,
J.C., Irigoin, H., Keller, M., Pomposiello, M.C. 1982. Resultats de
deux campagnes magnktotelluriques faites dans le Nord la
PBninsule Antarctique en 1979 et 1980. IX e Reun. Ann. Sci. de la
Terre, Paris, SOC. Geol. Fr. ... dit., Paris.
Fournier, H., Keller, M., Demichelli, J., Irigoin, H. 1980. Prospeccion
magnetotelurica en la isla Vicecomodoro Marambio, Antartida.
Direc. Nac. del Antartico, Inst. Antart. Argentino, Buenos Aires,
Contribucion 235: 38-43.
Fukuda, M., Strelin, J., Shimokawa, K., Takahashi, N., Sone, T.,
Trombotto, D. 1992. Permafrost occurrence of Seymour Island
and James Ross Island, Antarctic peninsula region. In: Y. Yoshida
et al. (eds.), Recent Progress in Antarctic Earth Science: 745-750.
12

Living microorganisms in Siberian permafrost and gas emission at low
temperatures
A. Brouchkov, M .Fukuda, K. Asano, M. Tanaka and F. Tomifa
Hokkaido University, Sapporo, Japan
Permafrost is a source of greenhouse gases. Thus thawing
of the frozen soils affects conditions of the global carbon cycle.
Organic material in the thawing permafrost decays
quickly, releasing carbon dioxide and methane in the process.
Long-term soil incubation experiments in flasks containing
soil samples and nitrogen in the air have shown a slow pre
duction of methane in different frozen soils at -5°C. Soils
wereover-saturatedwithwater.Therewasanincrease in
methane content in the air of the flasks, especially in the first
20-50 days of the experiments (Brouchkov and Fukuda
2002). The change in methane content occurred according b
the logarithmical law insamplesofmodemsoilsfromYakutsk,
Russia and Hokkaido (Tomakomai), Japan; the rates
of methane production decrease with time (Table 1).
world for many years, they might be still active. Techniques
for sampling in permafrost for microbiological studies are not
well-established. The requirement is for biological cleanliness
and absolute avoidance of contamination while the sample is
brought to the surface and during subsequent handling of the
sample. Fieldwork done in 2001-2002 in Yakutsk, Eastern
Siberia, has shown that the permafrost at temperature -5°C of
contains living fungi identified as Penidlium echinulafum
(Figure 1).
Table 1. Change of methane content (ppmv) in the air of the flasks
during an incubation experiment at -5%
Days Tomakomai soil Yakutsk soil
0 1.09 0.85
22 1.80 2.00
50 1.89 2.07
390 3.15 3.10
Micro-organisms could be responsible for microbial methane
formation in permafrost: according to recent studies,
methane and gas hydrates are widely distributed in there and
appear to be of a biogenic origin. Colonies of microorganisms
were discovered that survived the incubation pe
riod lasting longer than 1 year -5OCc; at they were able to undergo
anaerobic growth at room temperature in GYP me
dium; 16s rDNA was amplified by PCR.
However, calculations of the amount of methane prcduced
in permafrost based on short-term experiments are difficult.
The production could be discontinued under certain conditions.
The temperature is not stable: it could be suggested
that methane production increases at higher temperatures. The
type of soil, organic material and water content are also essential
to microbial activity. In spite of the obvious possibility
of methane production at low temperatures, long-term forecast
of methane content in frozen soils is problematic.
Permafrost soils contain micro-organisms (Friedmann
1994, Gilichinsky and Wagener 1995); isolated from the
The genera Penicillium are widespread and found in soil.
Fungi are responsible for the biodegradation of organic material;
this could be considered as one of their important practical
applications. Some fungal species grow at low temperatures,
below 0°C. However, there is no information
concerning the fungi which may grow in underground perma-
frost at temperatures below 0°C during a long time throughout
their life cycle.
The identification of permafrost fungi as Penici//ium
13
echinulafum was based on morphological characteristics and
a nucleotide sequence analysis of enzymatically amplified

Brouchkov, A, et al, : Microorganisms in Siberian permafrost, gas emission ...
18s rDNA, internal transcribed spacer (ITS) region including
5.8s rDNA and DllD2 at the 5' end of the large subunit (26s)
rDNA. The fungi were able to grow at positive and negative
temperatures. The optimum temperature for the growth of P.
echjnulaturn was estimated as I5-2OoC, yet they were able
to grow at the negative temperature of -5°C (Figure 1). Fungi
colonies on MEA grow rather rapidly, attaining a diameter d
20 to 35 mm in 7 to 9 days at 20°C.
organisms to those of known soil micro-organisms could re
veal this mechanism.
REFERENCES
Ashcroft, F. 2000. Life at the Extremes. HarperCollins.
Brouchkov, A.V. and Fukuda, M. 2002. Preliminary Measurements
on Methane Content in Permafrost, Central Yakutia, and some
Experimental Data. Permafrost and Periglacial Processes 3:
187-197.
Brouchkov, A.V. and Williams, P.J. 2002. Could microorganisms in
permafrost hold the secret of immortality What does it mean
In: Contaminants in Freezing Ground. Collected Proceedings of
2nd International Conference. Cambridge, England, Part 1: 49-
56.
Friedmann, E. I. 1994. Permafrost as microbial habitat. In: D. A Gilichinsky
(ed.): Viable Microorganisms in Permafrost. Russian
Academy of Sciences, Pushchino, Russia: 21-26.
Gilichinsky, D. and Wagener, S. 1995. Microbial Life in Permafrost:
A Historical Review. Permafrost and Periglacial Processes 6:
243-250.
Williams, P.J. and Smith, M.W. 1989. The Frozen Earth, Cambridge:
Cambridge University Press.
Sampling of permafrost exposures in Aldan river valley
(Mammoth Mountain, Eastem Siberia), in oneoftheoldest
regions of permafrost on Earth (possibly aged about three million
years) was done. Blocks of frozen soil and ice sized ap
proximately 20x20~20 cm in size were cut from exposures
and taken frozen to Sapporo. Frozen wood of the same age
was also sampled.
Micro-organisms have been discovered (Figure 2). They
were able to grow at both aerobic and anaerobic conditions.
For the sequence analysis DNA was extracted using the
ISOPLANT set (Nippon Gene) and following the manufacturer's
instructions. Preliminary results of sequencing and
analysis of the evolutionary tree of the micro-organisms show
that they are most related to Bacillus anthracis.
Understanding these' permafrost organisms and their relationship
to their cold environment, especially to unfrozen water
common in frozen soils (Williams and Smith 1989), has
immediate practical implications. The fact that viable bacteria
occurring at significant depths in permafrost can be prompted
to reproduce, presents imaginative possibilities, for example
for the decontamination of buried contaminants. The living micro-organisms
in permafrost, in common with other extremophiles
(Ashcroft 2000), apparently have special mechanisms
of repair of cell structures, necessary for their survival
(Brouchkov and Williams 2002). A comparative study d
structure and biochemical features of permafrost's micro-
14

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Natural risk evaluation of geocryological hazards (the Varandey Peninsula
coast of the Barents Sea)
1. V. Chekhina and F. M. Rivkin
industrial and Research lnstifute for Engineering of Construction, Moscow, Russia
INTRODUCTION
It is known that the intense technological impact on the Russian
Arctic coast caused by industrial engineering results in
the intensification of natural and origination of new geocryological
processes. Therefore, prediction of development d
geocryological processes is part and parcel of geotechnical
survey.
The intensity of geocryological processes can be estimated
only on the basis of complex engineering geocryological
studies. Therefore, the creation of information databases
on the basis of engineering geocryological studies is an element
of primary importance for solving the complex of prob
lems of environmental management.
In this connection, it becomes possible and necessary b
automate construction of different engineering geocryological
maps using geoinformation systems (GIs). This makes it
possible not only to solve traditional problems of engineering
survey but also to map natural risks and to assess the possibility
of appearance of geocryological processes and their intensity.
METHODS AND RESULTS.
the engineering geocryological zoning, using the number d
Results of the engineering geocryological studies performed exogenous processes and the degree of appearance risk d
on the Varandey Peninsula were used as base information b geocryological hazards. Thefirstevaluationtablehasbeen
estimate the intensity of geocryological processes. compiled using the total number of available hazardous ex-
The electronic database of geocryological conditions was ogenous processes (frost heaving, soil settlement upon
compiled on the basis of these studies. The database includes thawing, thermal erosion, landslides, and cryogenic appearsuch
attributes as engineering geological regions, which were ances) and ignoring the appearance risk in the regions distindistinguished
taking into account the geomorphological posi- guished. The second evaluation table reflects the total appeartion,
genesis and lithology of soils. Moreover, physical ance risk of the processes in each of the region, additionally
(moisture, ice content, density, etc.) and thermal (heat capac- including estimation of ice content and salinity of soils. The
ity, thermal conductivity, etc.) properties of thawed and frozen total score of complexity of each geocryological region is
soils were also taken into account. Each region distinguished
was indexed depending on geocryological conditions.
The database has a two-level architecture. Data obtained
as a result of the field studies and laboratory determination d
soil properties is combined at the first level. The second level
includes generalized information, namely first-level data and
information obtained from publications. Since the second-level
database includes information from different sources, all input
parameters are tested for mutual consistency.
As a result, the compiled (summarized) database provides
the complete set of properties necessary for analyzing
the environmental quality in each of the distinguished regions.
The structure of the obtained database is correlated with the
scheme of zoning and is factually the legend to the map.
Natural risks of such geocryological hazards as frost
heaving, soil settlement upon thawing, thermal erosion, landslides,
and cryogenic appearances were evaluated based on
the compiled database. Ice content and salinity of soils were
also estimated in order to assess the complexity of the area.
The rule base (analytic relationships for scores) and pro.
gram module for calculating scores were created in order b
evaluate natural risks of geocryological hazards. The module
for calculating scores from analytic relationships makes it
possible to determine an appearance risk of exogenous gemryological
processes (Dan'ko, 1982; Lovchuk, 1990). A
certain score is assigned to each cryogenic processes depending
on the obtained index value. The score system is
based on risk evaluations of each geocryological process
(Armand, 1975).
Thus, two evaluation tables were obtained. These tables
make it possible to differentiate regions, distinguished during
given for the most hazardous process, assuming that effect its
will be higher than the effect of the remaining processes.
As a result, all engineering geocryological regions were
differentiated with respect to the degree of the appearance risk
of hazardous processes in the following way: most
hazardous, medium hazardous, and least hazardous. Data d
both summary tables were reduced to the electronic engin-
15

eering geocryological map, which is directly related to the
database.
The electronic database of the key site of the Varandey Peninsula,
which includes the engineering geological characterization
of the natural environment, was created as a result d
the performed studies. The algorithm was constructed for
evaluating natural risks of appearance of such geocryological
hazards as frost heaving, soil settlement upon thawing, thermal
erosion, landslides, and cryogenic appearances and for
estimating ice content and salinity of soils.
Two evaluation maps of the key site of the Varandey Peninsula
were constructed on the basis of theelaboratedtechnique:
- Map of zoning with respect to the appearance risk of the
most hazardous geocryological process (Fig. 1);
- Map of distribution of exogenous geocryological processes
(Fig. 2).
CONCLUSION
The system for interpreting results of geotechnical survey,
using electronic maps, electronic databases, and program
modules, has been created as a result of the performed studies
in order to evaluate the distribution of exogenous processes.
Theelaborated system can be used to map natural
risks and to construct prediction maps of development of hazardous
exogenous processes.
ACKNOWLEDGEMENT
This work is supported by Russian Coast Protection Program
and INTAS grant 2332.
REFERENCES
Armand, D.L. 1975. Landscape Science, Moscow: Mysl'
Dan'ko, V.K. 1982. Regularities of Development of Thermal Erosion
in Northwestern Siberia, Cand. Sci. (Geol.-Min.) Dissertation,
Moscow.
Lovchuk, V.V. 1990. Cryomorphogenesis of Subarctic Lowlands in
Western Siberia: Backgrounds of a Topography Cryomorphological
Analysis, Methods of Engineering Geocryological
Survey, Moscow: VSEGINGEO.
Figure 1. Map of zoning of the key site of the Varandey Peninsula
with respect to the appearance risk of the most hazardous geocry-
ological process.
1 - most hazardous: 2 - medium hazardous; 3 I
least hazardous.
Figure 2. Map of distribution of exogenous geocryological processes
at the key site of the Varandey Peninsula.
1 - most hazardous; 2 - medium hazardous; 3 - least hazardous.
16

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available information
AIGEO working group “Past distribution of the periglacial environment in
the Italian mountain chains: reconstructions based on the study of relict
forms”
Alessandro Chelli, *M. Pappalardo and A. Ribolini
Dipartimenfo di Scienze della Terra, Universifa di Parma, Parma, lfaly
* Dipartimenfo di Scienze della Terra, Universifa di Pisa, Pisa, lfaly
A working group within the Italian Association of Geomor- periglacial landforms in their relict state (e.g., rock glaciers
phologists started its activity at the end of 2002. A number and wedge casts) for Quaternary paleoclimatic reconof
Italian researchers, engaged in studies on the perigla- structions in Italian low altitude mountain environments.
cia1 environment and often dealing with relict forms, decided
to get together in order to actively compare their
d 15 I
7
study objects and methods.
In Italy, landforms related to periglacial processes have
been recognized for a long time (Capello 1960) and permafrost
distribution in the present and throughout the
Quaternary has recently been reconstructed by means of
typical landforms (e.g. rock glaciers; Dramis and Gug-
/
lielmin 1999). However, many other landforms, e.g., different
types of inactive scree, common mainly in the Apennine
chain and in Sardinia, were identified in the past
and considered as “periglacial”. Although displaying different
characteristics with regard to parent material, texture,
form, position and extension, they have some fea-
1
tures in common: they are located on moderately steep
slopes, are made of a coarse debris that may be mixed
with fine material and they have no sedimentary features
typical of fluvial or gravitational landforms; in all cases
they are not related to present-day geomorphic processes.
In the literature they were classified mainly as eboulis ordonnees,
greze lite6e and block streamslfields.
The working group members intend to collect evidence
i
in order to clarify the true genetic environment of such
landforms: whether they are really periglacial, or whether Fig, 1 Location of the study areas.
just cryogenic weathering and frost shattering can be responsible
for coarse sediment supply, in the absence of Study areas were established (Fig. 1) and a prelimieven
sporadic permafrost. The possibility that other nary illustration of the deposits concerned permitted the
weathering processes have been responsible for the for- members to compare the different typologies. Within the
mation of such deposits, and that the action of gravity or category of block streamslfields study cases were serunning
waters in the transport of weathered material has lected in Sardinia, Tuscany and Western Liguria. Sites
until now been misunderstood, will be carefully examined. with slope deposits of likely cold climate environment were
In fact, stratified scree is no longer strictly considered a enclosed both from the Northern Apennines (Parma and
typical periglacial landform, as multiple processes can be Reggio Emilia sectors) and from the Central (Abruzzo and
evoked for its genesis (Van Steijn et al. 1995). Moreover, Molise) and Southern Apennines (Cilento), as well as from
block streams, presently being studied as active forms in Sardinia. Active landforms testifying to the present-day
extreme environments (very cold and arid), were recog- transition from glacial to periglacial environment in the
nized to be convergent, in their relict form, with other, non Gran Sasso area (Central Apennines), especially lobated
periglacial, types of landforms (Boelhouvers et al. 2002). debris indicating permafrost creeping processes, will be
Another aim of the working group is to employ typical included among the investigation objects. Their analysis
17

can be important, in fact, for making a comparison with the
studied relict forms.
The first goal of the working group will be the creation
of a work sheet in order to perform a description of the
different deposits with identical methods. Micromorphological
techniques and image analysis, successfully
employed by some of the members on single deposits, will
be extended wherever possible. Further work will focus on
the chronological aspect of the matter and the necessity of
finding a suitable dating method that permits comparison
of the ages of the studied bodies.
After four years’ activity, the goal of the working group
is to define the role that frost weathering had in sediment
supply, the role of frost creep and gelifluction processes in
the material accumulation, and the relationship between
the climate under which such landforms were created and
the distribution of mountain permafrost.
Working Group Members: A. Bini (Universita di Milano), N.
Casarosa (Provincia di Pisa), A. Chelli (Universifa di
Parma), M. Coltorfi (Universifa di Siena), P. Farabollini
(Universifa di Camerino), M. Firpo (Universiti di Eenova),
M. Guglielmin (ARPA Lombardia), E. Miccadei (Universifa
di Chieti), M. Pappalardo (Universifa di Pisa), M. Pecci
(ISPESL Roma), T. Piacenfini (Universitd di Chiefi), F.
Scarciglia (Universifi della Calabria), C. Queirolo (Universifa
di Genova), R. Rafi (Universita di Roma “La Sapienza’J,
A. Ribolini (Universifb di Pisa), s. Sias (Universita di
Sassari), C. Tellini (Universifa di Parma).
The abstract volume of the Working Group firsf meeting is
available upon request.
Chelii, A. et al.: AIGEO, periglacial environment. reconstructions, relict form
REFERENCES
Boelhouvers J., Holness S., Meiklejohn I. and Sumner P. 2002. Observations
on a blockstream in the vicinity of Sani Pass, Lesotho
highlands, Southern Africa. Permafrost and Periglacial Processes
13(4): 251 -257.
Capello C.F. 1960. Terminologia e sistematica dei fenomeni dovuti al
gelo discontinuo. G. Chiappichelli, Torino.
Dramis F. and Guglielmin M. 1999. Permafrost investigations in the
Italian mountains: the state of the art. In Paepe R. & Melnikov V.
(eds.), Permafrost response on economic development, environmental
security and natural resources, Kluver Ac. Publishers:
259-273
Van Steijn H., Bertran P., Francou E., HBtu B. and Texier J.P. 1995.
Models for the genetic and environmental interpretation of stratified
slope deposits: review. Permafrost and Periglacial Processes
6: 125-146.
18

8th ~ n ~ ~ Conference ~ ~ ~ on Permafrost ~ ~ i Extended ~ n ~ Abstracts l on Current Research and Newly Available i ~ f ~ r ~ ~ ~ i o ~
Active layer dynamics in Greenland, Svalbard and Sweden
H. H. Christiansen, *J. H. Akerman and **J. Repelewska-Pekalowa
lnstitute of Geography, University of Oslo, Norway
* lnstitute of Physical Geography and Ecosystems Analysis, University of Lund, Sweden
"*Department of Geomorphology, University of Maria Curie-Sklodowska, Poland
INTRODUCTION
Four sites with the longest active layer monitoring records
in high arctic NE Greenland and Svalbard, and in alpine
Sweden are compared. In the Greenland Sea, deep-water
formation occurs bringing warm surface water from the
south. This provides a relatively mild climate at high latitudes.
Meteorological data and GCMs all suggest that climatic
changes at high latitudes are amplified as compared
to lower latitudes. Therefore, active layer monitoring and
its relationship to climatic variations in Greenland, Svalbard
and northern Scandinavia are of high relevance for
monitoring and study.
The active layer monitoring sites are located at Abisko,
northern Sweden, 68ON 19OE, at Calypsostranda, 77ON
15OE, and at Kapp Linne, 78ON 13OE, both at Svalbard,
and at Zackenberg, 74ON 2I0W, NE Greenland. We discuss
intersite and interannual variations in active layer
thickness and its potential relationship to summer air temperature
represented by thawing degree days.
This abstract was developed as a result of the CALM
synthesis workshop held in November 2002 at the University
of Delaware, USA.
ground moraine at 36 m a.s.1 (Christiansen 1999). All sites
are located in open flat landscapes with average landscape
snow cover thickness and duration conditions. Meteorological
stations operated since grid establishment at
the Zackenberg Station 25 m from the Zackenberg grid;
until 1976 and again from 1996 at lsfjord 0.5-5 km from
the Kapp Linne grids; at the Hornsund Station, which is 68
km south of Calypsostranda; and at the Abisko Station 3-
50 km from the Abisko grids. In the last 20-25 years there
has been an increase (1-2 OC) in summer temperatures
from the Abisko and Longyearbyen, Svalbard stations.
This warming has, however, not occurred in Greenland.
DATA AND DISCUSSION
SITE DESCRIPTIONS
At Abisko, data on active layer thickness were obtained
beginning in 1978 at eight valley palsa bog grids at 380-
480 m a.s.l., and at three intermontane depression grids at
850-950 m a.s.l., each with 121 measuring points (Brown
et al., 2000). At Kapp Linne, monitoring was initiated in
1972 at ten I00 x I00 m grids on raised marine terraces
at 4-59 m a.s.l., nine on dry mineral soils and one in a
peat-covered area, each with 25 randomly sampled
measuring points until 1994, and after that with 121 points
(Brown et al. 2000). At Calypsostranda, active layer
monitoring started in 1986 at 23 individual points on raised
marine terraces at 2-40 m a.s.1. These points adequately
represent the average landscape active layer thickness
based on a total of 300 measuring points (Repelewska-
Pekalowa 2002). At Zackenberg, a grid of I00 x 100 m,
with 121 measuring points was established in 1996 on
1975 1980 1986 1990 21896
m
Figure 1. Active layer thickness (cm) at Abisko in Sweden, Kapp Lime
and Calypsostranda at Svalbard and at Zackenberg in NE Greenland.
No data was collected from the Calypsostranda sites in 1994, 1997
and 1999.
At the Abisko valley site, active layer thickness was
stable around 55 cm from 1978 to 1995, and increased to
around 80 cm since then (Fig. 1). In the Abisko Mountains
the active layer thickness varied from 85 to 120 cm during
the measuring period, with the largest values in the last
five years. The Kapp Linne sites show decreasing active
layer thickness until 1987 and then increasing, with the dry
sites varying around 75 to 110 cm, and the wet sites from
19

25 to 95 cm. The thickest active layer in the region occurs
at Calypsostranda, varying only from 115 to 140 cm for
both wet and dry sites. There is no trend in active layer
thickness at Calypsostranda. At Zackenberg the active
layer is only 60 to 70 cm thick. This is the smallest value in
mineral soil, and this site, with its relatively short record,
also has the smallest interannual variation of the four
sites.
A common pattern of interannual active layer thickness
variation exists from 1997 to 2000 (Fig. 1). The Calypsostranda
sites were not monitored during all these years, but
data from the monitored years fit the pattern. In 1980 and
1988 the Kapp Linne and Abisko sites have peaks in the
active layer thickness, but this was not seen at Calypsostranda
in 1988. In 1989 the active layer thickness was reduced
for all Kapp Linne and Abisko sites, but thickest
during the recording period at Calypsostranda, while in
1990 all Kapp Linne and Abisko sites peaked, when Calypsostranda
sites were small.
= 0 (n=14)
Calypsostranda wet
rz = 0.04 (n=14)
v v
V
v It 0.31 (n=30)
f ~ app Linne wet
P 0.59 (n=29)
Christiansen, H.H. et al,: Active layer dynamics in Greenland, Svalbard ...
Abisko mountain
P = 0.47 (n=26)
140
120
1 DO
shallow Zackenberg active layer in mineral soil of 60-70
cm is comparable to the peat sites in Svalbard and northern
Sweden. The shallow Greenlandic active layer reflects
the East Greenland Current carrying arctic water masses
south along East Greenland, causing relatively colder
summers here than at Svalbard and in northern Scandinavia.
On Svalbard, Calypsostranda is located only 58 km
south of Kapp Linne, and both sites are located on raised
marine terraces of the west coast. However, the active
layer at the Calypsostranda sites is significantly thicker
than even at the Kapp Linne mineral sites. The Hornsund
and Kapp Linne summer air temperatures are within the
same range, have equal overall interannual variations, but
with the Hornsund values slightly lower than the Kapp
Linne, particularly since 1995. During the active layer
monitoring period, summer air temperatures have been
rising mainly since 1998 at Kapp Linne, but not as significantly
at Hornsund. Whereas the active layer increased at
Kapp Linne already from 1988, there is no increase in active
layer thickness at Calypsostranda. This combination
of information indicates that local conditions at both sites
have a larger control on active layer thickness than summer
air temperature. This is in agreement with what has
been found for other CALM sites (Brown et al. 2000).
None of the locations have a significant correlation between
active layer thickness and DDT, stressing that the
control by local factors are as strong as the influence from
summer air temperature. It is also possible that other meteorological
conditions could influence the active layer
thickness as much as the summer air temperature. This
remains to be investigated. The interannual active layer
thickness variation pattern from 1997 to 2000 for all sites
show similar response, presumably controlled by some
regional effect, however, not by the air temperature alone.
REFERENCES
Figure 2. Active layer depth and the square root of thawing degree
days of the air temperature at the closest meteorological stations at
the four sites.
To investigate the relationship between active layer
thickness and summer air temperature, thawing degree
days (DDT) were calculated using the closest meteorological
station for each four investigated sites in the years
with active layer measurements (Fig. 2). There is a positive
correlation for the Kapp Linne and Abisko sites, and
particularly for the peat bog sites, but none for the Zackenberg
and Calypsostranda sites.
The main intersite difference in active layer thickness
between Greenland, Svalbard and Sweden primarily reflects
climatic and material differences. Both peat sites at
Abisko and at Kapp Linne have a generally shallow, 20-70
cm active layer, while the mineral soils here have thicker
active layers of 80-120 cm. The active layer in the peat
site at Kapp Linne is thinner than the Abisko alpine peat
sites, while the active layer in the Abisko Mountains is
slightly thicker than at the mineral sites at Kapp Linne. The
20
Brown, J., Hinkel. K.M., and Nelson, F.E. 2000. The Circumpolar Active
Layer Monitoring (CALM) Program: Research Designs and
Initial Results. Polar Geography, 24: 165-258.
Christiansen, H.H. 1999. Active layer monitoring in two Greenlandic
permafrost areas: Zackenberg and Disko Island. Danish Journal
of Geography, 99: 117-l2lI
Repelewska-Pekalowa, J. 2002. CALM - Site P1- Calypsostranda.7.
Akerman, J.H. 1980. Studies on Periglacial Geomorphology in West
Spitsbergen. Meddelanden frAn Lunds Universitets Geografiska
Institution Avhandlingar LXXXIX.

8th International Conference an ~~~~~~~~~~
Extended Abstracts an Current Research and Newly Available Information
Could the current warming endanger the status of the frozen ground
regions of Eurasia
S. M. Chudinova, S.S. Bykhovets, V.A. Sorokovikov, D.A. Gilichinsky, *T-J. Zhang and R. G. Barry
Soil Cryology laboratory, lnsfitute of Physicochemical and Biological Problems in Soil Science, Russian Academy of Sciences,
Pushchino, Moskow oblast, Russia
*Natinal Snow and Ice Data Center, University of Colorado, Boulder, USA
INTRODUCTION
The current rise in air temperature that started in the second
part of the 1960s in the northern territory of Russia
takes place predominantly in the cold period (Watson et
al. 1998). However, a number of impacts induced by climate
warming, such as reduction of the permafrost area
and changes in thickness of the active layer, etc., are determined
mainly by changes in ground temperature during
the warm period. Changes in ground temperature reflect
not only changes in air temperature but also combined
impacts from many other climatic factors. The seasonal
trends of ground temperature do not necessarily correspond
with the trends observed for air temperature (Gilichinsky
et al. 1998). Zhang et al. (2001) found that abundant
rainfall under global warming may lead to a decrease
in summer ground temperature. Therefore, the actual
danger of an air temperature increase can be assessed
only on the basis of direct determination of changes in
ground temperature. This work investigates the vast regions
of Eurasia, where the current warming represents a
potential danger.
METHODS
The investigation was made for the period 1969-1990, for
which the most comprehensive data on ground temperature
are available. Linear trends were used to estimate
tendencies in the time series of ground and air temperature.
Analysis was performed for the mean annual ground
temperature, mean annual and monthly air temperature,
as well as ground temperature averaged over the summer
(June-August), winter (December-February), spring
(March-May), and autumn (September- November) periods.
Ground temperature was measured at depths of 40,
160 and 320 cm. Best-fit linear trends were calculated for
individual stations. Regions with different responses of
seasonal ground temperature to the air warming were
thereby delimited (Fig. 1). Then, time series of average
trends of ground temperature were calculated for each region
by the method of polygon area averaging. Trends of
ground temperature equal to 0.04"Clyear and above are
significant at the 0.05 level.
RESULTS
A correspondence between changes in the time series of
annual air and ground temperature has been identified
down to a depth of 320 cm for the territory under investigation.
Maximum trends of air temperature, and hence
maximum trends of annual ground warming down to 320
cm are observed for the territories with permafrost and
seasonally frozen ground in the West Siberian Plain (region
IV), the central Siberian plateau (region VI), and the
mountains of Transbaikalia (region VIII). Here, trends of
ground temperature at the three depths analyzed average
0.07-0.06, 0.05-0.08, 0.05-0.06"Clyear, respectively. Annual
ground warming is practically absent in the zones of
discontinuous permafrost in Pribaikal'e (region VII) (0.01-
0.03"Clyear) and the seasonally frozen grounds of Far
East (region X) (O.O-O.O2"Clyear), and over vast territory
occupied by continuous permafrost to the east of the Lena
River (region XI) (-0.01-0.O"Clyear). This reflects an absence
of any significant increase in the annual air temperature
time series in these regions.
The relationship between mean seasonal time series of
air and ground temperature is more complicated. In the
taiga zone of the Russian Plain (regions I and II), changes
in ground temperature are controlled mainly by air temperatures
of the warm period (Chudinova et al. 2001). In
the northern and middle taiga zones (region I), in spite of
the predominant increase in winter air temperature, the
major rise in ground temperature is observed in the warm
period (Fig. lb), which is dictated by the strong rise in air.
The absence of summer air warming in the southern taiga
(region II) results in the absence of a ground temperature
increase during the whole year.
21

In the Ural mountains (region Ill) and West Siberian
Plain, a strong rise in air temperature in both the warm
and the cold periods results in large positive trends in both
winter and summer (0.08-0.1 O'Clyear) ground temperature
(0.07-0.10 in the upper layers and 0.04-0.08'Clyear
in the lowest one). In the vast permafrost zone of East Siberia
(regions VI, VII, VIII, and XI), trends in the ground
temperature time series are determined mainly by
changes in winter air temperature. The effect of changes
in summer air temperature extends only down to 40 cm.
Therefore, in accordance with air temperatures trends in
this region, ground temperature increases in the cold period
(autumn-winter-spring) and remains unchanged, or
decreases, in summer (Fig. 1). The largest significant rise
in ground temperature in the cold period (winter-spring),
down to a depth of 320 cm, characterizes the central Siberian
Plateau and mountains of Transbaikalia, averaging
0.10-0.08"Clyear. The summer trend of ground temperature
in these regions is zero or negative to a depth of 40
cm. Below, due to winter warming, trends of ground temperature
become positive and reach 0.05-0.06"Clyear.
Trends at individual stations can differ strongly from the
average estimates. Therefore, we attempted to ascertain
whether this increase takes place at stations where permafrost
exists from a depth of 160-320 cm. Our analysis
shows the absence of significant trends in mean monthly
ground temperature, both at the boundary of the thaw
layer and in the permafrost. The maximum decrease in
summer ground temperature is observed over the vast
discontinuous permafrost zone in region XI. A strong rise
in air temperature in February and March cannot offset
this cooling with depth. Thus, changes in permafrost
status in this region have not occurred.
P
-0.03 0.0 0.03 0.06 0.09
Chudinova, S. M. et at.: Current warming, frozen ground, Eurasia..
CONCLUSION
Despite a significant rise in winter and annual ground
temperature in the vast territory of East Siberia, no significant
effect on the permafrost status is apparent. Based on
the results obtained, attention should be focused on the
West Siberian Plain, where extensive territories are occupied
by permafrost and peatland. Changes in the temperature
regime of these ecosystems can have both
negative and positive consequences (for example, for agriculture).
The West Siberia Plain is the region of Russia
most sensitive to global air change (Anisimov 2001).
Therefore, changes observed in the ecosystems (especially
peatlands) of West Siberia may be of interest for
other regions of the world where similar changes are taking
place.
ACKNOWLEDGEMENT
This work was supported by the Russia Foundation for
Basic Research (02-05-64597) and in part by NSF-OPP
(9907541& 0229766).
REFERENCES
Anisimov, O.A. 2001. Predicting patterns of near-surface air temperature
using empirical data. Climatic Change 50: 297-315.
Chudinova, S.M., Bykhovets, S.S., Fedorov-Davydov, D.G., et al.
2001. Response of temperature regime of Russian North to climate
changes during second half of 20th century. Earth
Cryosphere 5(3): 63-69 (in Russian).
Watson, R.T. Zinyowera, M.C. and Moss, R.H. (eds). 1998. The Regional
Impacts of Climate Change. An Assessment of Vulnerability.
IPCC, Cambridge University Press.
Zhang, T, Barry R.G., Gilichinsky, D.A., Bykhovets, S.S., et al. 2001.
An amplified signal of climatic change in ground temperatures
during the last century at Irkutsk, Russia. Climatic Change 49: 41-
76.
Gilichinsky, D.A., Barry, R.G., Bykhovets, S.S. et al. 1998. A century
of temperature observations of soil climate: methods of analysis
and long-term trends. In A.G. Lewkowicz & M. Allard (eds), Permafrost;
Proc. the 7th intern. conf., Yellowknife, 23-27 June, 1998,
Universitk Laval: 313-17.
Figure 1. Regions (I-XI) with different linear trends of ground temperature
at a depth of 40 cm for (a) winter and (b) summer.

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Snow effects on recent shifts (1998-2002) in mean annual
temperature at alpine permafrost sites in the western
R. Delaloye and M. Monbaron
ground surface
Swiss Alps
Department of Geosciences, Geography, University of Fribourg, Switzerland
INTRODUCTION
The ground temperatures at several alpine permafrost sites
have been monitored for the last five years. The data analysis
highlights the effects of both duration and thickness of the
snow cover on the mean annual ground surface temperature
(MAGST).
The calculation of MAGST is frequently used for estimating
the occurrence of permafrost. Nevertheless, as weather
conditions may dramatically fluctuate from one year to another,
MAGST is susceptible to large interannual changes,
which can lead to significant misinterpretation of the data. The
snow conditions represent a key factor governing the shortterm
variations in MAGST. Concerning the amplitude d
these variations, Vonder Miihll et al. (1998) relate, for instance,
a MAGST drop of more than 2°C at a depth of 0.6 m
in a borehole over a one-year interval. However, the spatial
variability of the temporal evolution of MAGST still remains
poorly documented. The present work reports observations
made at both regional and local scales.
REGIONAL SCALE
Figure 1 provides a regional overview of the behaviour of
MAGST. For each site, the values correspond to the average
of all the UTL-I . A first observation is that the MAGST is
relatively high for permafrost terrain, frequently above 0°C.
This could be see as an effect of climate warming, with Ute
potential for a rapid degradation of permafrost. More probably,
however, these high temperatures indicate a "thermal offser
between the surface and the top of the perennially frozen
ground, a matter hat still requires further study of mountain
permafrost.
I I I I
the
SITES AND METHOD
The analyzed data stem from five sites, each less than 1 km*
-2.0
wide, and located within a radius of 20 km in the western ! I I I I I
1 1.38 l,l.9Y 1 1.30 1.1 31 l.l:oz Date
Swiss Alps. The regional climatic conditions are transitional
fd.rn.YY)
-R chy-N-2800-(4) -Mllle-NE-2350-(7)
between the humid northwestern fapde of the Alps (Mont- ..-... Ritord-W-2853-(17) -Aget-SE-2850-(1)
Blanc), the southern quite mediterrannean side of the alpine -.-...- Ssnetsch-N.2350-r71
range, and the drier inner parts of Valais. The Sanetsch site is Figure 1. Monthly running MAGST during the period 1998-2002.
an exception as it is directly located on the northern and wet- For each site, the values correspond to the average of n UTL-1 and
are placed at the end date of the annual interval. Legend: siteter
border of the Alps.
aspect-mean elevation-n. Black and white dots indicate the extreme
Depending on the site, 4 to 22 minidataloggers (UTL-I) values of MAGST calculated for all sites together, the arrows show
have recorded the temperatureat2-hour intervals, the first the date when they occurred.
ones since October 1997. The annual series are adjusted on
the zero curtain phases. The resulting accuracy of measure- Figure 1 also shows that the interannual evolution of the
ment is +/-0.15"C. The UTL-1 have been positioned immedi- MAGST looks quite similar for every site. An increase in the
ately below the ground surface in order to be protected from values from the end of 1999 until April 2001 was followed by
direct solar radiation, i.e., at a depth of between IOand 70 a strong decrease during the next 12 months. On average,
cm, depending on the ground granulometry and on the defini- over the 41 available UTL-1, the drop in MAGST reached
tion of the surface.
1.4"C between April 2001 and April 2002. The maximum
decrease obtained for a single datalogger was 3.I0C, the
23

minimum only 0.3%. The shift in ground surface temperature
was not accompanied by any significant change in mean annual
air temperature (MAAT, Fig. 1). Indeed, interannual differences
in snow covering caused this thermal drop.
Important snowfalls occurred relatively early in autumn
2000 and then protected the ground from winter cooling. The
effect on MAGST was an increase towards the maximal
values recorded during the five-year monitoring period. The
spring of 2001 was again characterized by subsequent
snowfalls which caused a relative delay in the snowmelt in
comparison with the previous year. In response, MAGST
began to decrease. The autumn of 2001 was not snowy and
was followed by very cold weather in December and early
January. Which resulted in especially cold ground temperatures.
In April 2002, MAGST at all sites reached their lowest
levels in the last 5 years at least.
A detailed comparison of the different sites reveals some
peculiarities which can be attnbuted to local differences in interannual
snow conditions.
LOCAL SCALE
Delaloye, R. et al.: Snow effects on recent shifts.
CONCLUSIONS
The analyzed data illustrate the interannual influence of snow
conditions on the annual ground temperature. MAGST can
vary sharply both in time and space without any change in
MAAT.
lnterannual variations in snow cover conditions at a regional
scale can induce strong changes in the ground surfac
temperature, which can be as high as I .5"C for
MAGST calculated over an entire area. Locally, or in mountainous
regions subject to more dramatic interannual shifts in
weather conditions than in the western Swiss Alps, interannual
changes in MAGST could be expected to be much
higher.
Interannual differences in predominant winds can result
locally in completely different snow cover distribution which
in turn has an influence the on spatial pattern of MAGST.
The above conclusions indicate in particular that singleyear
measurement of the MAGST in high mountain areas
must be very carefully interpreted.
The example illustrated on Figure 2 shows the effect of interannual
changes in snow conditions on the MAGST at a very
local scale.
The four UTL-1 involved were located on a complex of
push-moraines. In contrast with boththeprecedingand the
following years, the winter of 2000101 was characterized by
predominantly southerly winds which drastically modified the
snow distribution in the area. As a result, the spatial pattern d
the MAGST was inversted during 2001. The coldest places
at the end of 2000 became the warmest a few months later,
before being the coldest again in 2002. While the contrasted
thermal evolution at Ri-19 and Ri-20 can be easily explained
by their location on opposite sides of a pass, the behaviour of
Ri-21 and Ri-22 appears to be due to the presence of
plurimetric-sized ridges and furrows at that place. Variations
in the redistribution of snow by wind over a very short distance
- a few metres - is seen as being responsible for this
peculiar thermal behaviour.
REFERENCE
Vonder Muhll, D., Stucki, T. and Haeberli, W. 1998. Borehole
Temperatures in Alpine Permafrost : A Ten-Year Series.
Proceedings of the 7Ih International Conference on Permafrost,
Yellowknife, Canada. 1089-1095.
1 .0
h
0.0
E
2
-1.0
V'I -2.0
2000 2001 2002
Figure 2. Monthly running MAGST near the Six Rouge pass at the
Ritord site. On the location map, the white area corresponds to
loose sediments (mainly push-moraines). Outcropping bedrock is
drawn in grey. The Six Rouge is the small summit west of Ri-19.
24

Dangerous engineering-cryogenic processes at work in the urban
territories of Russia’s permafrost zone
E. A. Domnikova
Moscow State University, Faculty of Geography, Moscow, Russia
INTRODUCTION
Vast territories in Russia are located in permafrost zones.
Large natural resources are concentrated there and consequently
many buildings and facilities have been constructed.
Perennially frozen grounds are unstable with
temperature fluctuations and readily change their condition
from frozen to thawed and vice versa. These
changes are followed by a dangerous intensification of
geocryogenic processes with negative consequences for
the geoecological situation.
Cryogenic processes are geological processes which
are related to the perennial and seasonal freezing of
rocks (Kudjavtsev 1978).
The following main groups of natural cryogenic processes,
differing in their directions of influence and their
specific manifestations in Far North landscapes, can be
singled out: processes connected with warming effects
on permafrost (thermokarst, thermoerosion); processes
caused by cooling influences on the ground such as
thawed-ground freezing or additional ice segregation in
frozen ground (frost heave, cryogenic cracking); specific
cryogenic processes (slope processes, the process of
cryogenic weathering and the formation of frozen
grounds under specific salt conditions).
Natural cryogenic processes increase with industrial
development in permafrost zones and often new engineering-cryogenic
processes arise. In a lot of cases these
engineering-cryogenic processes produce disastrously
negative influences on the environment, buildings and
structures.
All engineering-cryogenic processes can be divided
into two major types: those having a natural analog and
their counterparts that have no complete natural analog.
The first type includes activated natural cryogenic processes
and also those caused by anthropogenic activity but
also occurring under natural conditions. The second type
embraces processes that are not found in nature.
To classify the processes belonging to the first type we
can use the groups cited for natural cryogenic processes
but adjusted for technogenic impact. Given below is a
classification of engineering-cryogenic processes that
have no natural analog since these processes are most
widespread in industrially developed territories of the permafrost
zone.
Technogenic heating of the ground is a process that
increases the temperature of the perennially frozen rock.
This process is intensified in industrially developed territories
if the needs of local environment are violated by the
maintenance of the facilities.
Technogenic salting of the ground is the process
whereby pollutants coming into the active layer alter the
properties of the ground. Frequently, under such conditions
of deep thawing within the boundaries of cities and
settlements cryopegs arise - unfrozen strongly mineralized
horizons with below zero temperatures. Technogenic
salting results in the active destruction of the foundation
material (Grebenets et a1.1997).
TECHNOGENIC INUNDATION
Most frequently, technogenic inundation is caused by
leaks from engineering structures. This process weakens
the ground structure and lowers its strength. It often involves
ground ice which leads to thermokarst and thus
deforming structures and foundations (Fig.1).

DESTRUCTION OF CONCRETE BY FROST
Domnikova, EA: Dangerous enginee~ing-cryogenic processes at work
The concrete of foundation pillars and supports is slowly
destroyed under repeated freezing - thawing conditions of
the moisture contained in the pores, microcrevices and
caverns (Grebenets 1995). When the water contained in
the cracks is converted into ice it expands and widens the
crack that is undergoing the freezing. This major factor,
producing negative effects on the foundation material, is
increased as the depth of thawing increases due to
elevated ground temperatures caused by technogenic
sources of heating.
It would seem advisable to classify not only the engineering-cryogenic
processes but also the degree of their
impact. This whole set of factors, leading to increasing
engineering-cryogenic processes, can be conditionally divided
into two groups: passive effects on the perennially
frozen rocks and active effects on the perennially frozen
rocks. Thus, all the engineering-cryogenic processes can
be divided into two groups: engineering-cryogenic processes
that only emerge under active effects on perennially
frozen rocks and engineering-cryogenic processes that
emerge even under passive effects on perennially frozen
rocks.
Various engineering-cryogenic processes occur with
different intensities under the same conditions of technogenic
effect. It would seem relevant to single out four different
categories of intensity:
I .Engineering-cryogenic processes with no marked
effects;
2.Engineering-cryogenic processes producing low intensity
effects;
3,Engineering-cryogenic producing high intensity effects;
4.Engineering-cryogenic processes producing disastrous
effects;
Engineering-cryogenic processes can be controlled by
stopping or neutralizing the processes of the technogenic
impact, or by taking special engineering measures so that
the geosystem reverts to its natural state, or by building
new properties to the requirements of the geoecology and
geocryology. Once this has become possible,
engineering-cryogenic processes are considered to be
controllable. If not, they are uncontrollable.
Studies of the dynamics of dangerous engineeringcryogenic
processes, during the industrial development of
the area have revealed their great dependence on regional
geocryologic peculiarities as well as on the character
and intensity of the technogenic impacts applied.
The map "The Intensification of Engineering-Cryogenic
Processes on the Urban Territories of Russia's Permafrost
Zone" (I:I
5'000'000) has been created for the first time. It
has been based on the "Geocryological Map of the USSR
and the following parameters were taken into consideration:
The character of the permafrost rocks in the area
(solid. unsolid); whether it is flat or mountainous territory;
the permafrost ground temperature and the level of zero
annual amplitude; the population of an inhabited area; the
26
type of industrial development and the predominant engineering-cryogenic
processes.
The problems of stable developments in permafrost
zones are very important now.
REFERENCES
Kudrjavtsev, V.A. 1978. General Geocryology. Moscow State University
Publisheryouse (in Russian).
Grebenets, V.I., Kerimov, A.G.-o. and Savtchenko, V.A. 1997. Stability
of foundation in cryolitic zone under the condition of technogenic
inundation and salting. Publications Committee of XIV
ICSMFE (ed). Proc. XlVth Int. Conference on Soil Mechanics and
Foundation Engineering, Hamburg, 6 - 12 September 1997: 113-
114. Rotterdam: Balkerna.
Grebenets V.I. 1995. Heat regime operation of frozen grounds on urban
areas. New constructions of foundaments and methods of
foundament preparation [Novyye konstruktsii fundamentov I metody
podgotovki osnovaniy]. VNllOSP Proc., VOI. 95. Moscow: 66-
79 (in Russian).

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Environmental maps of the Russian Arctic as a base for the segmentation
of the coastline and the analyses of coastal dynamics
D. S. Drozdov and Yu. V. Korostelev
Earth Cryosphere Institute, SB RAS, Russia
During the past years coastal dynamics in the Arctic Seas
have been studied under different aspects by various scientific
divisions and organizations. The main task of the
project is to investigate fundamental scientific questions
concerning coastal dynamics in the Eurasian Arctic by
using Geographic Information System (GIS) technology -
mapping, zoning and monitoring of the caostal exogenic
processes, based on geosystem principles.
Modern conditions and dynamics of the natural geosystem
in northern regions with permafrost areas exist
and constantly change, influenced by natural factors and
anthropogenic activities. To study the influence of these
factors, several GIS have been developed which can be
applied to produce cartographical models of the main
components of the geosystem. Digital environmental
maps at global, regional and local scales are the most
significant form of spatial realization of these models.
Various prognostic, eco-geological and environmental
maps were created and are being further developed at the
present time.
The GIS technology can be used to handle a geocryological
data base correlated with particular polygons on
the map and to create digital landscape, engineeringgeocryological,
eco-geological, environmental, etc. maps.
The GIS technology also supports monitoring needs, especially
in the case of significant human activity. If a forecasting
algorithm is accepted based on designed scripts of
natural dynamics and economic activity, the GIS and
monitoring system may represent not only modern data,
but also prognostic information on environmental parameters.
To analyze the dynamics of the Russian Arctic coastal
zone a set of digital landscape and Quaternary geology
maps with scales of 1:4 000 000 - 1 :2 500 000 for the hard
copy are currently being developed. The first draft of these
maps (Fig.1) was published within the framework of the
CAVM project (Drozdov et al. 2001, Walker et al. 2002).
According to the recommendations of the Arctic Coastal
Dynamics (ACD) project (Rachold et al. 2002) and the
Russian World Ocean project, the mapped areas will be
enlarged and more details on landscape and geological
units will be provided, especially along the coastline. Several
climatic zones from forest-tundra to central taiga will
be included in the mapped area and new types of landscape
and geological units with different lithology and
permafrost extent appear on the map. The consequent re-
processing of the source and derived maps, of satellite
imagery and of the data base have to pass through several
iterations.
The basic information which is needed to create the
environmental-geological GIS is collected in small-scale
graphic documents presented in the literature and in archives.
The most important information can be taken from
maps published by I.S.Gudilin in 1980, G.S.Ganeshin in
1976, J.Brown and others in 1997 and E.S.Melnikov and
others in 1999. Additionally, the results of own research
and the processing of remote sensing information were
used.
Figure 1. The landscape map of the Russian Arctic coastal zone: Climatic zonal, subzonal and altitudal-longtitudal landscape units.
27

The major part of the information needed for the
analyses of coastal dynamics can be extracted from the
landscape map of the Russian Arctic coastal zone which
provides a subdivision of landscape unit types. This
principle of unit types is widely used for drawing up
landscape, geocryological and environmental-geological
maps of the northern territories of Russia. It can be applied
for all scales, the only differences are the dimensions
of the examined and displayed landscape and
geological units. In the case under discussion, case we
deal with the units of largest dimension. Each landscape
- forming attribute affects on-shore exogenic processes
including coastal dynamics. The main landscape forming
controls are the following:
Location of the area in the particular natural-climatic
zone or subzone (e.g., arctic tundra, northern foresttundra,
central taiga, etc.).
Location of the territory in units of altitude zoning
(i.e., plains, plateau, mountains).
Genesis and main morphological attributes of relief
(e.g., landscapes of marine plains and terraces, glacial
landscapes, erosive-denudation landscapes of
mountains and piedmonts, etc.).
Characteristics of sediments: lithology of soils and
petrography of rocks (e.g., clay, sand, debris material,
karst-able and non-karstable bedrocks, etc.).
Each of the listed attributes has to be mapped and
the overlay of all these maps represents the final landscape
map, which serves as the base for spatial characteristics
of modern geological processes. Formally, all
lines on the obtained landscape map can be estimated
as the boundaries of common range. For convenience of
analysis, however, it is useful to organize the boundaries
in the order mentioned for the landscape-forming controls
where the boundaries of natural-climatic zones are
the most important and the boundaries of ground types
are least important.
The landscape and surface geology maps obtained
serve as a graphical base to make segmentation of the
Russian Arctic coastlines for description and analysis of
coastal processes and for further assessment of coastal
dynamics (Fig. 2).
At the present time, segmentation of one coastal zone
is practically completed. Each segment is to be supplied
with a table of appropriative data according to the appropriate
legend - the ACD classification template. This
work is currently being performed by specialists from
different institutes and organizations involved in the ACD
project.
ACKNOWLEDGEMENTS
This study is supported by INTAS (grant 01-2332) and
RFFl (grant 01-05-64256).
REFERENCES
Circum-Arctic map of permafrost and ground ice conditions. 1997.
Edited by J. Brown, O.J. Fenians, Jr., J.A. Heginbottom, E.S.
Melnikov - IPA. US Geological Survey.
Drozdov D.S., Ananjeva (Malkova) G.V. 2001. Landscape Map of
the Russian Arctic. I/ Forth Int. Circumpolar Arctic Veg. Map.
Work: Moscow, April 10-12: 43-45.
Melnikov E.S. (ed.) et alp, 1999: Landscape map of Russia permafrost
at scale 1.4 000 000. Tyumen. Earth Cryosphere Institute
S0 RAS (in Russian).
Rachold, V., Brown, J., Solomon, S. 2002. Arctic Coastal Dynamics
- Report of an International Workshop, Potsdam (Germany) 26-
30 November 2001, Reports on Polar Research 413, 103 pp.
Walker D.A. et al. 2002. The circumpolar Arctic vegetation map. -
Remote Sensing, V.23, #21: 4551-4570.
Figure. 2. The segmentation of the Russian Arctic coastal zone based on landscape and surface geology maps.
28

8th International Gonfcrence on Permafrost Extended Abstracts on Current Research and Newly Available Information
Use of BTS measurements and logistic regression to map permafrost
probability in complex mountainous terrain, Wolf Creek, Yukon Territory,
Canada
M. Ednie and A.G. Lewkowicr
Department of Geography, University of Ottawa, Canada
Many thousands of square kilometers of north-western and 2002. Miniature data loggers installed in fall 2000
Canada are underlain by mountain permafrost. At present, showed that stable minimum temperatures at the
even the most up-to-date permafrost maps of this area are snow-ground interface were reached at both times of
at scales which classify whole regions as sporadic or measurement, as required for the BTS methodology.
widespread discontinuous permafrost, thereby ignoring 2) A map of potential incoming solar radiation for the
the influence of elevation and exposure. In order to pro- basin was developed using a 30 m digital elevation
vide baseline data for assessment of the effects of future model and the Solar Analyst extension (Fu and Rich
climate change and to assist with mineral development, a 1999) in Arcview 3.2. The proportion of incoming rasimple
methodology is needed to measure and identify diation modeled as direct beam was set using cloud
permafrost distribution in this complex mountainous ter- cover data from Whitehorse.
rain. The BTS (basal temperature of snow) method, widely 3) Measured BTS values were regressed against elevaemployed
in the European Alps (e.g. Hoelzle et al. 1993) tion and the modeled potential incoming solar radiaand
elsewhere, could be of great use in this context since tion and gave a statistically significant relationship
it allows the development of a model of permafrost likeli- with an r*of 0.37. This value is comparable to those
hood within a GIs. However, a thick, stable snow cover (at from European investigations (Gruber and Hoelzle
least 80 cm deep) is required to insulate the ground from 2001). The scatter in the graph of measured vs.
atmospheric temperature variations in order to use BTS modeled BTS values (Fig. 1) likely derives from varitemperatures
as indicators of permafrost. Studies of the ables that are not included in the modeling, such as
thermal influence of the relatively thin and wind-blown albedo and vegetation cover.
snow cover that is typical of northern Canada (e.g. Gran- 4) We determined the relationship between modeled
berg 1973) have led to skepticism among some research- BTS and permafrost distribution, rather than employers
as to the potential efficacy of the BTS method in this ing the categories used in the European Alps. In late
region. Consequently, to our knowledge, this is the first July-August 2002, holes were augered or pits were
attempt to test the BTS method to map mountain perma- excavated at a total of 200 sites with a range of elefrost
in North America.
vations and slope orientations to verify the presence
Our study was carried out in the Wolf Creek research or absence of frozen ground. Frozen ground enbasin,
20-30 km south of Whitehorse in an area mapped countered within 2 m of the surface, or predicted from
as part of the sporadic discontinuous permafrost zone. instantaneous temperature profile measurements to
The basin has an area of almost 200 km2 at elevations of be present in this layer, was taken to be an indication
750 to 2250 m a.s.1. and a vegetation cover that changes of permafrost.
from boreal forest to alpine tundra as elevations increase. 5) Logistic regression (e.9. Pereira and ltami 1991) was
The annual air temperature at Whitehorse is -0.7OC used to evaluate the relation between the modeled
(1971-200) at 703 m a.s.1. and a value of -4OC was meas- BTS values and the permafrost distribution identified
ured from 2001-02 at an elevation of 1235 m within the in the excavations. A cluster of points from a confined
basin.
valley in the central part of the basin were significant
The methods used to evaluate the efficacy of the BTS outliers and indicated that permafrost was present
methodology to predict the probable distribution of perma- there at relatively low elevations. It is believed that
frost in this complex mountainous terrain were as follows: these resulted from winter air temperature inversions
which are common in the Wolf Creek basin. These
1) A geo-referenced set of 396 BTS measurements points were excluded from the analysis. Traditional
were collected within the basin in March-April 2001 BTS results (e.g. Hoelzle et al. 1993) are given as
29

-10 -8 -6 -4 -2 0
Measured BTS (“C)
Ednie, M. el aL: Use of BTS measurements and logistic regression ...
qualitative values (permafrost improbable, permafrost between BTS values and permafrost probability (Fig. 2)
possible, and permafrost probable) whereas logistic needs to be tested in other parts of the Yukon Territory.
regression allows for the calculation of quantitative
probabilistic values of permafrost distribution based
on BTS (Figure 2). For Wolf Creek, there is a >90% REFERENCES
probability of permafrost at modeled BTS values c-
5OC, and -I .I‘C. Analyst: Arcview an extension for modeling at so
scape scales. Accessible at:
A preliminary probability permafrost map Of Wolf
http://www.esri.com/librarvluserconf/~roc99/~roceed/~a~ers/pap8
Creek basin was produced (not shown). From the 671~867.htrn
analysis, about 40% of the basin is underlain by Per- Granberg, H.B. 1973: Indirect mapping of the snow cover for permamafrost.
frost prediction and In Proceedings Scheffew
Second International Conference on Permafrost, North American
Contribution, Yakutsk, USSR. Washinaton, - DC: National Academy
of Science Publication: 113-120.
Gruber, S. and Hoelzle, M. 2001: Statistical modeling of mountain
permafrost distribution: local calibration and incorporation of remotely
sensed data. Permafrost and Periglacial Processes 12: 69-
77.
Hoelzle, M., Haeberli, W. and Keller, F. 1993: Application of BTS
measurements for modeling mountain permafrost distribution. In
Proceedings of the Sixth International Conference on Permafrost,
Beijing. South China University of Technology, Beijing. Vol.1: 272-
277.
Pereira, J.M.C. and Itami, R.M. 1991: GIS-based habitat modeling
using logistic multiple regression: A study of the Mt. Graham red
squirrel. Photogrammetric Engineering & Remote Sensing 57(11):
1475-1486.
Figure I Measured BTS versus modeled BTS values.
1 r--
0.9 -
e
0.8 -
0.7 -
0.6
n
IC 0.5
0
-.-
0.4 -
n
n
n
0.1 -
0 4” ,
-11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
BTS (“C)
Figure 2: The probability of permafrost curve.
We conclude that BTS modeling to predict permafrost
is possible in mountainous parts of the Yukon Territory
and that a quantitative probabilistic approach is the most
logical given the variable terrain, vegetation and snow
cover in this basin and elsewhere. Future research is
needed to identify the topographic characteristics of
mountain valleys that may experience air temperature inversions
and to incorporate such characteristics into a
permafrost prediction model. In addition, the relationship

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Modeling the influence of moisture variations on physical rock weathering
processes in Antarctica: proposed investigations and preliminary results
c. Elliott
Geography Department, University of Canterburyl Christchurch, New Zealand
The primary objective of the proposed research is to
quantify the role of moisture in the weathering of rock in a
cold environment such as Antarctica. It is being conducted
to fulfill the requirements of a Ph.D. Although primarily focussed
on physical weathering processes, some investigations
of the potential role of chemical and biological
weathering will also take place. Specifically, the research
aims to develop a model to predict changes in weathering
rate, for a particular rock type, along a temperature and
moisture gradient on the Victoria Land Coast, Antarctica
(Figure IA). One of the more challenging aspects of investigations
into rock weathering is the complexity of factors
involved and Antarctica is an ideal environment to undertake
such research as its low moisture (at least in the
area under consideration here) and extremely low temperatures
enable some of these factors to be excluded or
minimized.
Despite the fact that the weathering of rock is a fundamental
landscape process, the important role that moisture
plays in weathering, although recognized, has not yet
been quantified. For example, there has been no systematic
study of how increasing quantities of moisture may influence
the type of process operating. Nor, apart from the
work of Prick (1997), have attempts to quantify the influence
of changing moisture availability on weathering rates
been made. With one or two notable exceptions (e.g.
Matsuoka 1991), very few rock weathering studies have
attempted to develop models that will predict weathering
rates. A series of simulations will be conducted on rock
samples from four locations in the field using an environmental
cabinet and the effects of moisture isolated from
other factors. The use of an environmental cabinet and
speeded-up climatic cycles is a well-established technique
for measuring weathering rates (e.g. Warke and Smith
1998). However, in this study the climatic cycles to be
used in the laboratory will be based on field observations
of temperature and moisture regimes within the rock,
rather than those using air temperature, for example. To
achieve this, water baths will be used to provide the
moisture levels necessary and the cabinet fitted with a
lamp as an artificial 'sun' to enable the rock surface temperatures
to be achieved.
An empirically-based mathematical model will then be
developed that will predict changes in weathering rate depending
on changes in moisture in the rock. The weathering
rate will be estimated by both measuring changes in
the weight of the samples and using a non-destructive
sonic technique to detect changes in strength (e.g. Fahey
and Gowan 1979), after a specific number of cycles.
Four locations along the Victoria Land Coast have
been identified to undertake the fieldwork: Victoria Valley,
Gneiss Point, Terra Nova Bay and Cape Hallett (Fig. IB).
These locations enable both a coast-to-inland perspective
(between Victoria Valley and Gneiss Point at appioximately
77.5 OS), as well as a 5' latitudinal gradient (between
Gneiss Point at 77.5 'S and Cape Hallett, at 72.5
O S ) to be investigated. Hourly recordings of surface and
subsurface (45 mm, 90 mm) rock temperature measurements
from mid-October to midJanuary will be made using
thermocouples and dataloggers. One-minute temperature
measurements will also be taken during each
visit (Le., in mid-October and mid-January) in order to determine
whether any thermal gradients likely to induce insolation
weathering have occurred. Year-round measurements
will be attempted during the second field season in
2003104.
Surface moisture sensors will provide measurements
of hourly surface moisture data, and relative humidity
probes will enable subsurface moisture regimes to be determined
(at 45 mm and 90 mm depths). The latter technique
has not yet been used to measure subsurface
moisture content in the field in rock weathering studies,
but has been used successfully for measuring moisture
content in concrete. These measurements are recorded
manually and so can only be undertaken during the times
of visits.
These techniques represent a unique aspect of this research,
as previous studies, such as those undertaken by
Hall (1986), have measured average moisture content of
rock samples left in the field rather than investigating levels
of moisture at different depths in the rock. In addition,
macro-climatological data (air temperature, relative humidity,
wind speed and direction) provided through existing
Automatic Weather Stations will be analyzed.
31

~
~~
~ ~
Table 1: Mean vapor densities at Gneiss Point October 2002 and
January 2003 at 45 mm and 90 mm depth in the rock
Depth of Vapor Density (gmlm2)
Probe October January
45 mm 1.4 5.1
90 1.2 mm
5.2
Figure 2. Vapor densities for the Gneiss Point Site at 45 mm and 90
mm depths in January 2003 indicate that the 90 mm vapor densities
are greater than the 45 mm vapor densities during the evening and
early morning, but are less than the 45 mm ones during the day.
REFERENCES
Island
Fahey, B. D. and Gowan, R.J. 1979. Application of the sonic test to
experimental freeze-thaw studies in geomorphic research. Arctic
and Alpine Research 11 (2): 253-260.
Hall, K. 1986. Rock moisture content in the field and laboratory and its
relationship to mechanical weathering studies. Earth Surface
Processes and Landforms 11 : 131-142.
Matsuoka, N. 1991. A model of the rate of frost shattering: application
to field data from Japan, Svalbard and Antarctica. Permafrost and
Periglacial Processes 2: 271-281,
Prick, A. 1997. Critical degree of saturation as a threshold moisture
level in frost weathering of limestones. Permafrost and Periglacial
Processes 8: 91-99.
Warke, P. A. and Smith, B.J. 1998. Effects of direct and indirect heating
on the validity of rock weathering simulation studies and durability
tests. Geomorphology 2: 347-357.
Figure 1. (A) General location of field sites in relation to the Antarctic
continent (6) The Victoria Land Coast indicating the four field sites at
Victoria Valley, Gneiss Point, Terra Nova Bay and Cape Hallett. Not
to scale.
The first fieldwork season has just been completed at
Gneiss Point and Victoria Valley and rock temperature
and moisture data successfully captured. Initial results
give a clear indication of the different temperature regimes
between the October and January recordings, as well as
the changes in moisture levels at the different depths
within the rock. Mean moisture levels between October
2002 and January 2003, although similar between depths
for each period, show significant differences between periods
(Tablel) and during the course of the day (Figure 2).
32

8th International Conference an Permafrost Extended Abstracts on Current Research and Newly Available l ~ ~
The lower discontinuous permafrost boundary in the Argentera Massif
(Maritime Alps, Italy): insight from rock glacier geo-electrical soundings
D. Fabre, *A. Ribolini, P. R. Federici and M. Pappalardo
Laboratoire lnterdisciplinaire de Recherche lmpliquant la Geologie et la Mecanique (LIRIGM), Universite Joseph Fourier, France
* Dipartimento di Scienze della Terra, University of Pisa, Pisa, Italy
The rock glaciers of the Argentera Massif are located in
the southernmost part of the European Alps, only 50 km
from the Mediterranean Sea. Glacial and periglacial reconstructions
in this Alpine region indicate out a rock
glacier formation mainly during the Younger Dryas and
Subboreal cold phases (Finsinger and Ribolini 2001 and
references therein). In this marginal area of the Alps,
transitional climate between Mediterranean and more
continental conditions made uncertain the evaluation of
rock glacier activity, that in some cases seems controlled
by local microtopography and meteorology
(Federici et al. 2000). Recently, Surface Ground Temperature
(SGT) of some rock glaciers (Ribolini 2001),
allowed to site the Lower Discontinuous Permafrost
Boundary (LDPB) at about 2,600 m a.s.l., which is
slightly lower than the one proposed for the southern
French Alps in the opposite flank of the mountain chain
(Evin and Fabre 1980).
To improve this evaluation, during the years 2001-02
geo-electrical soundings were conducted on four rock
glaciers (namely: Bassa Margot, Agnel, Portette and
Prefouns), selected at mean elevations ranging between
2,450 m and 2,700 m, thus proximal to the inferred
LDPB. All the analyzed rock glaciers belong to the
Gesso Basin in the southeastern Argentera Massif.
The Schlumberger method was adopted (with AB12
max ranging from 20 to 70 m), using an apparatus
(MAATEL BMI) working at DC 2000 V of max power.
This device was calibrated on the Murtel rock glacier in
Switzerland, where a mechanical sounding through the
permafrost is available (Fabre and Evin 1990). Previous
soundings in the southern French Alps (Assier et al.
1996 and references therein) evidenced a typical active
rock glacier vertical profile of resistivity: a) surface layer
without permafrost (5,000-20,000 Dm), b) ice-saturated
sediments (permafrost) with resistivity ranging from tens
of thousand up to million of Wm, and c) bedrock or not
frozen basal sediment (1000-10,000 nm).
Two longitudinal profiles in the upper-middle part
(BMI, BM2) and a transversal one (BM3) in the lower
part were performed on the Bassa Margot rock glacier.
The results (Fig. 1) indicate presence of permafrost with
medium-high ice content layers in the upper-middle part,
while in the lower one unfrozen sediments were found.
The Portette rock glacier also shows high resistivities.
In the upper part transversal profile (POI), data are
consistent with a 5 m thick permafrost level with a medium
ice content (110,000 Om), overlaying a 20 m thick
of permafrost with an higher ice content (-1 Mnm).
Similar resistivities in the vertical distribution results from
the transversal profile in the lower part (P02) (Fig. 1).
The three transversal profiles (PRI, PR2, PR3) performed
on the Prefouns rock glacier show the presence
of permafrost, with thick layers characterized by low to
medium ice contents. Here, the results of the soundings
are consistent with a permafrost thickness of 20 m
overlaid by a few metres of sediments with little ice
content (Fig. I).
Agnel rock glacier was investigated by means of only
one transversal profile (AG), because of its limited areal
extension. The resistivity distribution found (Fig. 1) suggests
a thin layer of permafrost (I to 7 m depth), with
low ice content (70,000 nm) probably not in equilibrium,
characterized by a temperature near the melting point (T
> -1 "C)
Considering that the rock glaciers studied could be
considered representative of all the rock glaciers that are
present in the Argentera, a speculation about activity
elevation can be proposed. Sectors of rock glacier developed
above 2,600 m are affected by permafrost, up to
20 m thick and with medium-high ice content (20 to 40
%). Rock glaciers at elevations ranging between 2,450
and 2,600 m are affected by thinlrelict permafrost layers.
Below 2,450 m it is unlikely to find permafrost, except for
very isolated, sheltered sites.
On the basis of these results, the Argentera Massif
can be considered the southernmost discontinuous permafrost
environment of the Alps, whose lower boundary
corresponds to about 2,600 m a.s.1.
33

Fabre, Q. et at, : Permafrost boundary, Argentera Massif, rack glacier, geo-electrical soundings.
I
REFERENCES
Fabre, D. and Evin, M. 1990. Prospection electrique des milieux a
tres forte resistivitk le cas du pergklisol alpin. Proceeding of
Sixth International Congress of the International Association of
Engineering Geology, Pays-Bas: 8.
Assier A,, Fabre, D. and Evin, M. 1996. Prospection Blectrique sur
les glaciers rocheux du cirque de Ste-Anne. Permafrost and
Periglacial Processes 7: 53-68
Finsinger, W. and Ribolini, A. 2001. Late glacial to Holocene deglaciation
of the Colle del Vei del Bouc-Colle del Sabbione Area
(Argentera Massif, Maritime Alps, Italy-France). Geografia Fisica
e Dinamica Quaternaria 24: 141-156.
Federici, P.R., Pappalardo, M. and Ribolini, A. 2000. On the ELA
and lower limit of discontinuous permafrost boundary in the
Maritime Alps (Italian side). Atti Accademia delle Scienze di Torino
134: 23-29.
Ribolini, A. 2001. Active and fossil rock glaciers in the Argentera
Massif (Maritime Alps): Surface ground temperatures and paleoclimatic
significance. Zeitschrift fur Gletscherkunde und Glazialgeologie
37: 125-140.
34

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available information
Using catchment soils to establish the provenance of high arctic pond
sediments: Truelove Lowland, Devon Island, Canada
LA. Fishback and *R.H. King
Scientific Coordinator, Churchill Northern Studies Centre, Canada
*Deparfment of Geography, The University of Western Ontario, London, ON, Canada N6A 5C2
INTRODUCTION
High arctic ecosystems have been identified as critically
sensitive areas in terms of global changes because of the
strong negative feedback mechanisms that create sensitivity
to the effects of greenhouse warming. Lakes and
ponds in these regions have been identified as “sensitive
bellweath-ers” of these changes (Douglas et al. 2000).
Ponds are defined as water bodies that freeze to the
bottom each winter whereas lakes maintain an unfrozen
water column during the winter months.
The preservation of the paleo-environmental record in
arc-tic pond sediments has been found to be more complete
than anticipated (Douglas et ai. 2000). However,
such paleo-environmental reconstructions tend to be far
more effective in determining past stages of the water
body than providing indi-cations of environmental changes
in the surrounding catch-ment areas.
This research attempts to address the catchment
condi-tions with respect to pond sediment records by
gaining a bet-ter understanding of the physical, chemical
and mineralogical signal of catchment processes and
identifying the presence of this signal in the adjacent high
Arctic pond sediments.
STUDY AREA
Truelove Lowland (75040’N, 84035’W) is one of a series
of coastal lowlands with high biological diversity located
on northeastern Devon Island, Nunavut, Canada. As a result
of isostatic emergence following the last glaciation, a
series of raised beaches was formed (King 1991). These
features are topographic barriers that have trapped water
resulting in the formation of approximately 530 ponds and
five lakes (ap-proximately 20% of the Lowland area) (King
1991). Pooh Pond in the central Lowland and the adjacent
raised beach were selected for this study.
METHODS
The sediments of Pooh Pond, a large, 1.2 m deep,
freshwa-ter pond were characterized using four short
sediment cores (e 0.40 m) taken from a variety of water
depths in a transect from shore to deepest area of the
pond. Sediments were cored using a modified corer (id =
51 mm) and fractioned at 50 mm intervals. Fractions were
analyzed using chemical fractionations and selective extractions
to characterize the sediments and differentiate
between materials from the catch-ment and the water
body, as well as establishing indicators of specific catchment
processes.
Major soil inspection pits were excavated in a catena
comprising four landscape units within the catchment. Soil
ho-rizons from each pit were sampled in duplicate in the
field and processed with the pond sediment cores. A data
set was compiled of similar biogeochemical variables for
pond sedi-ment and soils and tested for normalcy. A series
of multivari-ate statistics were used to investigate the
provenance of the pond sediments.
RESULTS AND DISCUSSION
A 14C pond sediment basal date of 5980 f 40 yrs BP
(Beta-162840) from 0.32 m depth was comparable to a
predicted emergence date for Pooh Pond, indicating a
freshwater origin of pond sediments collected. The lack of
biogeochemical stratigraphy in these pond sediment cores
was most likely due to cryoturbation or wave-induced turbation
of the sedi-ments that act to homogenize the unfrozen
sediments. An ANOVA of the physical and chemical
characteristics of the four cores demonstrated a significant
difference between the four cores. These differences indicate
the importance of local sedimentary environments in
the pond and the difficulty of single core characterization
of pond sediments.
The four soil types investigated on the raised beach
are formed on poorly stratified gravelly carbonate-rich
beach de-posits (Lev and King 1999). Soils on the raised
beach crest were poorly developed with desert pavement
at the surface and a thin A horizon overlying a coarse parent
material with carbonate pendants and particle caps.
Brunification dominates the soils in the upper and lower
foreslope as indicated by the Na-pyrophosphate extractable
Fe in the soils in Table 1. These soils are more developed
with high organic carbon contents, a thin Bm horizon
and an overall finer texture. Contemporary
translocation of Fe and AI in the lower foreslope soil soh-
35

tions was indicative of incipient podzoliza-tion. Soils at the
slope base are waterlogged with accumu-lated organic
matter over a gleyed silty parent material with high concentrations
of Mn and Fe.
Table 1: Extractable Fe as an indicator of brunification in soil horizons
in the raised beach soils.
Soil Twe 1 Soil Horizons
1 Feo 2
(Landscape Unit)
ReWosolic Static Cryosol Ahk
(% wt)
0.050
(Raised Beach Crest) Cca 0.025
C k2 0.020
Brunisolic Eutric Static Crvosol 0.080 Ah
(Upper Foreslope) 0.050 Bmk
Ckl 0.035
C k2 0.025
Brunisolic Eutric Turbic AhY Cryosol 0.110
Fishback, LA. et 81.: Using catchment soils to establish the provenanca..
Agri-Food Canada Publication 1646 (Revised).
Douglas, M.S.V., Srnol, J.P. and Blake, W.Jr. 2000. Summary of paleolimnological
investigation of high arctic ponds at Cape
Herschel, east-central Ellesmere Island, Nunavut. In M. Garneau
and B. Alt (eds.), Environmental Response to Climate Change in
(Lower Foreslope) BmY 0.1 15
Ckyl 0.025
Cky2 0.030
Gleysolic Turbic OfY Cryosol 0.210
(Meadow)
W Y 0.050
1 as determined by the Canadian Soil Classification System, 1998.
2 Na-pyrophosphate extractable Fe to represent organically chelated
Fe in the B horizon in % of soil weight.
Cluster analysis was used to classify the raised beach
soil horizons into four groups to create a reference set for
comparison with the pond sediments. The ‘unknown’ pond
sediments were classified into the soil groups using a discriminant
function analysis (DFA). The inorganic pond
sedi-ments were classified predominantly with soil horizons
showing some weathering. This DFA also identified
the most discriminating variables in classifying the pond
sediment with the soils as total organic carbon, organically
chelated Fe, AI and Mn and ‘plant available’ elements
(Mehlich Ill extract).
A second cluster analysis of the entire soil and pond
sediment data set resulted in separate groups of pond
sedi-ments and soil horizons with the exception of one
group. This group contained soil horizons from the toe of
the slope and from the pond sediment core closest to the
shore highlighting the localized connection of soils and
pond sediments.
chemistry of sediment alone. Pond sediments are a result
of a number of genetic processes operating together to
produce a unique material. Modification of the catchmentderived
material dur-ing transport and deposition obscures
the catchment signal and therefore influences the interpretation
of the sediment record.
REFERENCES
Agriculture and Agri-Food Canada, Soil Classification Working Group.
1998. The Canadian Soil Classification System. Agri-culture and
the Canadian High Arctic. Geological Survey of Canada Bulletin
529.
King, R.H. 1991. Paleolimnology of a polar oasis, Truelove Low-land,
Devon Island, NWT, Canada. Hydrobiologia 214: 317-325.
Lev, A. and King, R.H. 1999. Spatial variation and soil development
in a high arctic soil landscape: Truelove Lowland, Devon ls-land,
Nunavut, Canada. Permafrost and Periglacial Processes. 9: 35-
46.
CONCLUSIONS
Soil profile morphologies reflected a variety of pedogenic
and cryogenic processes within a soil catena across a
raised beach in Truelove Lowland. These processes included
brunification and incipient podzolization.Sediments
from the adjacent Pooh Pond were heterogeneous and
the characteri-zation of these sediments based on single
cores was difficult. Biogeochemical analyses indicate that
Pooh pond sediments are predominantly catchmentderived
in origin.
This research indicates that while these pond sediments
are derived mostly from catchment soil material, it
is difficult to establish provenance based on the biogeo-
36

~ I
~
8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Atmospheric and geothermal conditions during thermal contraction of ice
wedges, Bylot Island, eastern Canadian Arctic archipelago
D. Fortier and M. Allard
Centre d'ktudes nordiques, Universite Lava/, Quebec, EIK 7P4
We studied thermal contraction cracking of ice-wedge
polygons located on a terrace bordering a glacio-fluvial
outwash on Bylot Island (730 09'38"N, 800 0043,21 W).
Previous studies showed that the polygons of the study site
are composed of a syngenetic ice-rich accumulation of insitu
grown peat and wind-blown silts (Fortier and Allard
2001).
Air and ground temperature conditions preceding the
rupture of electric cables buried in the active layer over ice
wedges were analyzed for the period 1997 to 2002. Frost
cracking (breaking of electric cables 5 mm 0.d.) occurred
from November to March with the largest number of breaks
(45%) taking place in January, during the polar night. Due
to the absence of solar radiation, air temperature variations
are closely linked to the thermal properties of the air
masses. In the Bylot Island region and in northeastem
Baffin Island, the atmospheric depressions in the winter
come mainly from the south (Maxwell 1981). Analysis d
wind data recorded over the study site showed that frost
cracking of ice wedges occurred mainly following the
incursion and persistence of eastward and southward cold
weather accompanied by winds of varying strength from
the Byam Martin mountains and Navy Board Inlet.
The rate of geothemal contraction in the polygons is
related, in a non-linear way, to the length, range and speed
of the cooling of the air temperature, the thickness of the
snow cover, the sediment types and the ice content of the
ground (MacKay 1993, MacKay and Burn 2002). Crack
openings occur when contraction deformation by creep in
ice wedges is not fast enough to release the mechanical
tensions created by the thermal contraction of the terrain.
From 1997 to 2002, the mean daily air temperature on
days when cracking occurred was -36 "C, with air temperatures
ranging from -27 to -40 "C. Frost cracking of the
ground (electric cable ruptures) occurred a few hours to a
few days after a drop in the mean hourly air temperature in
the order of 7.9 "C over an average of eighteen hours at a
mean air cooling rate of -0.45 "C/h. Maximum cooling mte
was -0,9 "Clh when air temperature quickly dropped by
18.6 "C in less than 24 hours (12-2000). Fifteen of the
nineteen cracking events occurred when the mean daily air
temperature was stationary or rising slightly since the pre-
ceding day. This is due to the penetration time of the cold
wave into the snow cover and the frozen ground. Table 1
shows air temperature cooling rates during the days before
selected frost cracking events, for the 1997 to 2002 period.
Table 2 shows mean, maximum and minimum ground
temperature occurring during frost cracking days. During
cracking episodes, the mean temperature at the ground
surface (-2 cm) was -22.9 "C(I
997-2002) and ranged from
-14.7 "C to -29 "C. Geothermal data obtained from a
thermistor cable, installed in permafrost in 1999, reveal that
mean daily ground temperature at the top of permafrost (40
cm under the ground surface) during frost cracking events
(I 3) was -18.6 "C (2000-2002) and varied from -23.8 "C b
-12.8 "C.
1 YEAR I DAY / MONTH I ATCR ("C I h)
1997
1997
1997
21 - 22/02 (24)
13 14/03 (15)
14 - 15/03
-0,61(-8,5"C/14h)
-0,44 (-6,2"C/14h)
-0,43 (-6,9"C/l6h)
I ISM I IO- iil0l I -0.55 (-9.4"C/l7hl I
I999
I999
9
2000
2OOO
2000
22/01 (26)
23 - 24/02 (26,27)
24 - 25/02 (26,27)
21/01 (23)
23/01
(13)
-0,21 (4,9"C/23h)
-0,36 (-9,3"C/26h)
-0,58 (-7,O0C/I2h)
-0,53 (-5,8"C/l I h)
-0,69 (-6,2"C/09h)
-0'33 (-7,3"C/22h) 12/02
2001 (5)
-0,51 (-7,2"C/14h) 04/02
2001 26 - 27111 (28)
-0,36 (12,4"C /34h)
2001 27 28/11
-0,26 ( SOC I 19h)
37

Fortier, D. et al. : Atmospheric, conditions, thermal contraction, ice wedges, Bylot Island ...
Table 1. ATCR : air temperature cooling rate preceding selected
frost cracking events of the 1997 to 2002 period. Frost cracking
days in bold.
Lachenbruch, A.H. 1962. Mechanics of thermal contraction cracks
and ice wedge polygons in permafrost. Geological Society of
America, New-York: 69.
MacKay, J.R. 1993. Air temperature, snow cover, creep of frozen
ground, and the time of ice-wedge cracking, western Arctic
coast. Canadian Journal of Earth Sciences 30: 1720-1729.
MacKay, J.R. and C.R. Burn. 2002. The first 20 years (1978-1 979
to 1998-1999) of ice-wedge growth at the lllisarvik
experimental drained lake site, western Arctic coast, Canada.
Canadian Journal of Earth Sciences 39: 95-111.
Table 2. Mean, minimum and maximum daily ground
temperatures ("C) at given permafrost depths, during frost
cracking days (2000-2002).
Cracking episodes that occurred from 2000 to 2002 had a
mean thermal gradient in the active layer (0-40 cm) of 10.9
"Clm that varied from 3.9 to 18.8 "Clrn. Deeper down, the
mean thermal gradient was 3.34 "Clm (1.1 to 5.5 "Clm)
between 40 and100cm,3.12 "Clm (-2.1 to 4.3 "Clm)
between I and 2 m and 1.64 "Clm (0.7 to 3.25 "Clm)
between 2 and 3 m. According to our analysis of the mean
ground cooling rate occurring at different depths before frost
cracks openedin2000-2002,the maximum cooling zone
was primarily located in the active layer, most of the time
being in the upper part close to the surface (table 3). Between
0 and 20 cm, the mean daily groundcooling rate
was -0.25 "Clday and varied from -0,l to -1,l"Clday.
Since the tensile strength of the ice-rich peaty loess is
greater than the tensile strength of wedge ice, frost cracks
probably initiated at or near the top of permafrost, Le., the
top of the wedges. (Lachenbruch 1962).
DEPTH MDGCR
Surface -0.25
-0.21
Table 3. Mean daily ground cooling rate (MDGCCR; "Clday) at
given permafrost depths during frost cracking days (2000-2002).
REFERENCE
Fortier, D. and Allard, M. 2001. Cryostratigraphy and chronology
of syngenetic ice wedge polygons, Bylot Island, Canadian
Arctic archipelago. Proceedings 3Ist international Arctic
workshop, University of Massachussetts: 31-32,

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Results of a first study on tree growth and soil characteristics at a cold site
located below the limit of discontinuous alpine permafrost
M. Freppaz, L. Celi, *V. Stockli and M. Phillips
Swiss federal Institute for Snow and Avalanche Research, Davos Doe Switzerland and Laboratory Research Center on Snow and
Alpine Soils, Italy
“Swiss Federal lnsfjfufe for Snow and Avalanche Research, Davos Do$ Switzerland
INTRODUCTION
Scree slopes with particularly cold ground temperatures at
low elevations have been known for centuries in Switzerland.
They were termed ‘wind holes’, ‘ice holes’ or
‘weather holes’ by local farmers who recognized the potential
of such locations for cooling milk and butter during
the hot summer months. The sites are generally steep
scree slopes located at the foot of high cliffs.
Studies effected over the past 50 years have demonstrated
that the ground is not frozen as a result of a cold
microclimate at these locations, but that it occurs due to a
particular underground air circulation phenomenon and
due to the presence of an insulating vegetation cover
(Delaloye and Reynard 2001).
A particularly interesting and important characteristic is
that despite their occurrence in montane areas, these
sites have subalpine and alpine vegetation, and that some
plants, in particular spruce (Picea abies L. Karst) display
signs of severely limited growth. The soils generally reveal
a high organic matter content, with the formation of a humus
mor type at the surface.
In this preliminary study we analyzed the main properties
of these cold soils and of tree growth in order to
elaborate the first hypotheses about the contribution of soil
characteristics to the reduced plant growth.
METHODS
The Creux du Van in the Swiss Jura, at 1200 m a.s.l.,
where the patches of dwarf trees are especially well developed,
was selected as a test site. To compare soil
characteristics and tree growth in areas with different evidence
of permafrost occurrence the site was divided into
two strata, corresponding to the prevailing tree growth
form: I) stunted forest, and II) normal spruce forest.
In each of the two strata a spot was selected where a
soil profile was dug and tree cores were sampled. The
distance between the two sampling points, situated at the
same elevation and aspect, was about 50 m.
Around each soil profile in every stratum, three trees of
representative sizes were selected. The height of the trees
was estimated and the tree stems were cored perpendicular
to their axes at heights of 30 cm above the ground.
In the laboratory, the wood cores were processed with the
usual dendrochronological methods (Schweingruber
1988).
To record soil temperature, thermistors combined with
dataloggers (UTL-1) were buried in pairs in the two soil
profiles at a depth of about 30 cm (Oa horizon). The loggers
recorded the temperature throughout the period November
2001 to November 2002.
The soil pH was determined in a 1 :20 soil-water suspension.
Total soil carbon and nitrogen were measured
using a THERMOQUEST NC 2005 combustion analyzer.
The microbial biomass C and N were determined by the
fumigation technique. Humic acids (HA) and fulvic acids
(FA) were extracted from the soil and the elemental C and
N content determined. The N availability was determined
by anaerobic incubation of soil samples under water in
closed containers at 40°C for 7 days.
RESULTS AND DISCUSSION
The analysis of tree ring patterns revealed that the smallest
trees of stratum I exhibit ages around 90 years, despite
their small size (mean height = 3 m). Trees in stratum
II are distinctly larger (mean height = 25 m), older
(124 years) and have much more rapid growth rates.
Soil profiles typically exhibit an upper soil horizon (Oi)
of a fibric humus type, an intermediate horizon (Oe) of a
hemic humus type and a lower horizon comprising a
sapric humus type (Oa), which filled the interstices between
blocks of limestone.
According to the thickness of the active layer, estimated
at 2 m under the dwarf trees (Delaloye and
Reynard 2001), the soil was classified as gelic Histosol
(stratum I) (ISSS Working Group RB. 1998). The depth of
the active layer under the normal forest was estimated
39

greater and the soil was classified as sapric Histosol
(stratum 11).
In both strata the humus type in the Oi and Oe horizons
comprises mainly Sphagnum residues, which actively
acidify the soil upper horizons and hence are responsible
for the low pH (4.2 - 4.3). The pH in the Oa
horizon of all three strata is higher than in the upper horizon,
reaching values of 7.3 to 7.4.
Mean soil temperature (Oa horizon) between November
2001 and November 2002 (Fig. I) was higher in stratum
II (+3.43 "C) than in stratum I (t2.63 "C) as expected
by the depth of the active layer.
To test these hypotheses, and to better understand soil
chemical status and plant reaction, further investigations
are underway.
REFERENCES
Dalias, P., Anderson, J.M., Bother, P. and Couteaux, M.M. 2001.
Long-term effects of temperature on carbon mineralisation processes.
Soil Biology & Biochemistry 33: 1049-1057.
Delaloye, R. and Reynard, E. 2001. Les eboulis geles du Creux du
Van (Chaine du Jura, Suisse). Environnements pbriglaciaires (Association
Franqaise du Periglaciaire) 8: 118-129.
ISSS Working Group RB. 1998. World Reference Base for Soil Resources.
International Society of Soil Science (ISSS), International
Soil Reference and Information Centre (ISRIC) and Food and Agriculture
Organization of the United Nations (FAO). Report No.84,
Rome, 91.
Myrold, D. 1987. Relationship between microbial biomass nitrogen
and a nitrogen availability index. Soil Science Society of America
Journal 51: 1047-1049.
Schweingruber, F.H. 1988. Tree Rings: Basics and Application of
Dendrochronology. D. Reidel Publishing, Dordrecht.
-4 .J
200 1 2002
Figure 1. Soil temperature recorded in the two soils (Oa horizon)
(stratum I grey; stratum II black) from November 2001 to November
2002. Average values of the 2 loggers in each stratum.
Soil organic matter content is high in both strata, due to
the low soil temperature which may have contributed to
reducing the C decomposition rate. The Corg and Norg
concentrations of the Oa horizon were respectively equal
to 382 g C kg-I and 15.9 g N kg-I in stratum I and 428 g
C kg-I and 12.9 g N kg-I in stratum II. In the Oa horizon
of stratum I the C and N content of the humic fraction
(70.3 g C kg-I; 7.53 g N kg-I) was lower than in stratum II
(103.6 g C kg-I; 8.08 g N kg-I), revealing how temperature
might affect the biochemical pathways and decomposition
in addition to their metabolic rates (Dalias et al.
2001).
The microbial C and N pools are significantly lower in
stratum I (1.1 g C kg-1; 0.12 g N kg-I) than in stratum II
(17.5 g C kg-I ; 2.50 g N kg-I).
The N availability assessed by incubation under anaerobic
waterlogged conditions was equal to 29.2 mg N
kg-I in stratum II and 14.8 mg N kg-I in stratum I. The
amount of mineralizable N obtained in stratum II is in the
range of values previously determined for coniferous forest
soils (Myrold 1987) while in stratum I it is significantly
lower, revealing a limited N availability to plants.
From these preliminary results it appears that in stratum
I the low soil temperature recorded affects the soil
biological activity, influencing the biochemical pathways
and reducing the nutrient availability to plants. In stratum
II, less affected by frost conditions, the humification process
carried out by the microbial biomass is more pronounced
and influences the quality of the soil organic
matter. The N availability is higher and may contribute to
explaining the greater tree growth rate.
40

Spatial and temporal observations of seasonal thawing in the Northern
Kolyma lowlands
D. G. Fyodorov-Davydov, V. A. Sorokovikov, A. 1. Kholodov, V. E. Ostroumov, S. V. Gubin and D. A. Gilichinski,
*PI. S. Mergelov, **S. P. Davydov and S. A. Zimov
Soil Laboratory, Institute of Physicochemical and Biological Problems in Soil Science, Russian Academy of Sciences, Pushchino,
Moscow Region, Russia
*Soil Geography Laboratory, lnstifufe of Geography, Russian Academy of Sciences, Moscow, Russia
"'Northeast Science Station, Pacific Ocean institute of Geography, Russian Academy of Sciences, Far East Branch, Chersky,
Russia
Between 1996 and 2002, 15 sites (100x100 m, Fig. 1)
were established in the Kolyma lowlands within an area of
250x300 km for active-layer monitoring (Brown et al.
2000). Eleven of the sites represent different tundra subzones:
arctic (Kuropatochya River), typical (Cape Chukochiy,
Yakutskoe Lake) and southern tundra (Chukochya
River, Alaseya River, Konkovaya River, Segodnia pingo,
Ahmelo Lake). The sites differ also in lithology of parent
materials, which accounts for the differing appearance of
the landscape in various sectors of the region. Observations
started at some sites in 1989.
Figure I. Map of the measurement sites: R12A - Kuropatochya River
(top of the Edoma), R12B - Kuropatochya River (slope of the Edoma),
R13 - Cape Chukochiy (Edorna), R13A - Cape Chukochiy (alas), R14
- Chukochya River, R15A - Konkovaya River (alas), R15B - Konk-
ovaya River (Edoma), R16 Segodnya pingo, R17
- Akhmelo Chan-
nel, R18 - Mountain Rodinka, R19 - Glukhoe Lake, R20 - Malchikovskaya
Channel, R21 - Ahmelo Lake, R22 - Alazeya River, R25 -
Yakutskoe Lake.
Ice-rich sediments of the Edoma formation (loess-icy
complex) are widespread in these northwestern lowlands.
These deposits were subjected to intensive thermokarst
development during the Holocene optimum, which has resulted
in the creation of deep depressions or alases.
Thermokarst, alongside with hydraulic erosion, creates
high relief with considerable changes in elevation between
watersheds (edomas), alases and river terraces. Moundy
micorelief is common on watersheds and slopes. Vegetation
is represented by low shrub-grass-moss, grass-mossdryas
or cottongrass-moss-osier associations. The soil
cover consists of complexes with typical cryozems [Glacic
Haploturbels, Typic Haploturbels] and peaty-weakly gleyic
soils [Glacic Aquiturbels, Typic Aquiturbels]. In alas Valleys
or depressions, various polygonal bogs with peatyduff
[Typic Aquorthels], peaty-gleyic, peat-gley [Typic
Aquorthels] and peat [Sphagnic Fibristels, Typic Fibristels,
Typic Sapristels] soils are formed.
Ice-poor sand depositions of the Khallerchinskaya tundra
(Ahmelo Lake, Segodnia pingo) were also influenced
by thermokarst processes. Furthermore, they initially occupied
lower hypsometric levels than the Edoma surfaces.
All these processes resulted in considerable peneplaneization
of the surface, a general landscape waterlogging,
with an extremely high abundance of lakes and assoicated
migration. Because of the moundy microrelief, higher
landscape drainage is common in the southern part of the
area. Zonal biogeocenoses on sandy deposits have low
mound-hummocky microrelief, low shrub-lichen associations
and podzolized podbur [Spodic Psammoturbels] as
the zonal soil subtype. Polygonal bogs vegetation is represented
by moss, grass-moss and moss-lichen-grassy
associations, the soil cover consists of peat-duff or peatduff-gleyic
soils [Psammentic Aquorthels].
The northern taiga subzone is represented by Gluhoe
Lake and Mountain Rodinka. Low mound-hummocky
nanorelief, low shrub-lichen larch open forests and weakly
podzolized sandy podbur [Spodic Psammorthels, Spodic
Psammoturbels] are common for the first site; moss-shrub
larch open forests and gleyic loamy cryozem [Typic
Haploturbels] are formed at the second one.
The Ahmelo and Malchikovskaia Channels sites were
established on the floodplains biogeocenoses with high
heterogeneity. These sites cover both drained parts of
river terraces and different floodplain bogs. On river flood
terraces with moundy microrelief, calamagrostis associations
and sod-gleyic soils [Typic Umbriturbels] are developed.
Among floodplain bogs, sedge-cotton grassy plant
41

Fyodorov-Davydov, D. G. et a4.: Spatial and temporal observations of seasonal thawing.. .
associations with peat-gley soils [Typic Aquorthels] and
comarum-sedge associations with peaty-gley soils [Typic
Aquorthels] prevail.
The lithology of the parent sediments leads to significant
differences in the depth of active layer between the
north-east and north-west zonal tundra ecosystems of
Kolyma lowland: Dry sands thaw 2-3 times deeper than
the moist Edoma Ioams. The differences are considerable,
also for the zonal sandy soils (Ahmelo Lake) and soils of
polygonal bogs (Segodnia pingo) (Table 1).
soil thawed to a considerably shallower depth. This tendency
is better expressed for the zonal tundra sites (Cape
Chukochiy, Yakutskoe Lake, Alaseya River, Ahmelo Lake)
and northern taiga (Gluhoe Lake, Mt. Rodinka) biogeocenoses,
and least pronounced for intrazonal ones (Ahmelo
Channel, Segodnia pingo). A relation between active
layer depth and air temperature at the intrazonal sites of
Konkovaya River (alasj and Malchikovskaya Channel i S
not evident.
" - . . . . .-
Table 1. The mean depth of active layers for Kolyma lowland sites.
0-
Sites
Years
199 I 199 I 199 I 200 I 200 I 200
Figure 2. The 1989-2002 changes of the maximal active layer thickness
on different microrelief forms (sandy tundra soils at Ahmelo
Lake).
ACKNOWLEDEMENT
This report was developed as a result of the CALM synthesis
workshop held in November 2002 at the University
of Delaware.
*- top of the edoma
** - alas.
REFERENCES
In tundras of the western part of the Kolyma lowlands,
the climatic zonality is given at weekly intervals but still
leads to some increase in thaw depths from north to south
in soils of watersheds and slopes. In arctic tundras, the
mean long-term active-layer depth varied between 32 and
37 cm, in typical tundras between 36 and 38 cm, and in
southern tundras between 35 and 47 cm. The comparison
of two sites at Cape Chukochiy revealed greater thaw
depths on the upland watersheds than on waterlogged
alas lowlands. At the same time, in the basin of Konkovaya
River, warming effects could be due to a thicker
snow cover in winter on the alas than on the watershed.
The soils of floodplains, both at the boundary between
tundra and taiga (Ahmelo Channel) and in the northern
taiga (Malchikovskaia Channel) differ from zonal soils at
the same latitude with a considerably smaller active layer
depth.
The 1989-2002 data do not show any directional trends
in the changes of the active layer thickness (Fig. 2). However,
this relatively long time series indicates the dependence
of active layer depths on summer temperature. The
increase of thaw depth in the last three years (2000-02)
corresponds with high average summer (June-August)
temperatures (11.3-12.1oC) at most sites (Table 1). In the
years (1998-99) with lower temperatures (8.3-9.8oC) the
42
Brown, J., Hinkel K. M. and Nelson, F. E. 2000. The Circumpolar Active
Layer Monitoring (CALM) Program: Research Designs and
Initial Results. Polar Geography 24 (3): 165-258.

8th ~ n ~ Conference ~ ~ ~ on ~ Permafrost ~ i ~ Extended ~ a Abstracts l on Currsnl Research and Newly Available Information
Landscape indication of permafrost degradation in Central Siberia
S. P. Gorshkov and MA. Tishkova
Moscow State University, Moscow, Russia
One of the systems that is very responsive to climate
warming is the periphery of the permafrost zone, i.e., the
area of high temperature (- 0.10 -- I .OoC) permafrost. It is
important to know what is happening and what will happen
to the permafrost and permafrost landscapes in the coming
decades.
During the last 5-6 years the table of high-temperature
permafrost in the lower reaches of the Stony Tunguska
River has descended I m or deeper. Degradation takes
place on about 90% of the total permafrost area is in that
region.
According to the elaborated classification (Gorshkov, et
al., 2003) sensitive, or the least resistant, permafrost is
associated with the sites of low top surfaces (altitudes
about 200-250 m) and gentle slopes (steepness about 2-
30). The largest areas of such permafrost are situated in
the field of Pleistocene moraine and lacustrine-glacial deposits.
Both are represented by thin sands, aleurite, loam
and clays, sometimes with boulders. These dynamic permafrost
landscapes are swampy, with low sparse (canopy
4040%) cedar pine-spruce taiga forests with some
birches and larches, sometimes with firs, on peaty-gley
permafrost soils with moss-and-underbrush ground cover
and reindeer lichens in places. Solifluction is identified by
the presence of tilted trees (“drunken forest”) and gaps
(so-called “ruptures-windows”) of 1-2 meters in diameter
and 0.5-0.6 meters in depth, which disturb the turf-plant
layer. Usually gaps are filled with water during warm periods
and have a viscous-flow clay mass at the bottom. Almost
all above-mentioned landscape features are also
typical to the areas covered by birch forests. One important
difference is the distribution of trees. Two, three, four,
sometimes five or six and even seven-eight birches from 8
to 15 meters high grow together like big bushes. They alternate
with small wet lawns.
At the end of the summers of 1995-1997, the thickness
of the active layer was not more than 0.8-1.0 m in the
mentioned landscapes. But during the same period of
2002 it was about 1.2-1.5 m and sometimes 1.9 m and
deeper. The gaps were mostly without water. The rate of
active layer thickness increase is estimated to be 10-
15 cmlyear during 1998-2002. Sites of heat-deficient
slopes, glaces, and glacis floodplains are characterized by
the presence of more resistant permafrost. Under forest
vegetation at these sites we observed the permafrost table
at a depth of 0.5-0.8 m. We suppose that the active layer
thickness increased not more than 0.2-0.3 m during last
four years. More pronounced changes occurred in areas
located near thermokarst ponds and where forests or
dense bush groves are absent. In the first case the per-
mafrost table is located at a depth of I .O and I .3 m under-
neath holes and hummocks. In the second case geocryological
changes have not been so significant.
Three types of permafrost sites are the most resistant.
These include kurums and ((hanging)) peat bogs of heatdeficient
slopes, as well as forested floodplains of major
rivers.
Figure 1. Overgrowing kurum edge on the left slope of the Rybnaya
River valley in its middle reaches
Kurums could probably be transferred to the previous
category (Fig. I). An increasing number of blocks in unstable
positions is typical for kurums. Furthermore, many
blocks and spaces between them are being covered with
lichens. Water, which flowed at the base of the block horizon
has disappeared. The cold air between the blocks became
warmer. lt seems that during the last several years
a lowering of the permafrost table took place at kurum
43

~~
~~
"
Gorshkov, S.P. et al.: Landscape
sites. Perhaps ice disappeared from the base of kurums in
some places.
The geocryological situation remains without apparent
changes at hanging bog sites. The position of the permafrost
table is approximately at the former level: about 0.4-
0.5 meter under holes and 0.7-0.8 m under hummocks.
Permafrost is most stable at sites of boggy forested
flood-plains in the largest river valleys of the region. This
was proved by the investigations carried out in the valleys
of Kulinna, Stolbovaya, Vorogovka and Stony Tunguska
rivers (Fig 2). At these sites the depth of permafrost table
is about 0.2-0.3 m under holes and reaches 0.4-0.5 m
under hummocks.
indication of permafrost ~ ~ ~ .. ~ a ~ a ~ i ~ ~ "
8) signs of noticeable warming thoroughly of underground
kurum space after water supplying decrease;
9) possible lowering of the permafrost table in kurum
stows;
IO) intensification of thermokarst processes;
11) wide development of long-frozen rocks which can
indicate recent permafrost degradation in places;
12) occurrence of young fir-tree stands within burnt areas
previously covered with different coniferous
species.
During the last few years, a considerable increase in
active layer depth has occurred in the permafrost landscapes
of Central Siberia. Degradation takes place on
about 90% of the permafrost area in the region.
Intensification of solifluction and thermokarst is associated
with this process as well as transformation of kurums.
Degradation of high temperature permafrost is possible in
places. However, several types of sites which are most
resistant to global warming do not show clear responses
to current climate change. Nevertheless the abovementioned
processes, particularly solifluction, can be considered
as "trigger" mechanisms representing a largescale
response to global warming in the studied region.
ACKNOWLEDGEMENT
Figure 2. Floodplain under birch taiga on peaty-boggy permafrost soils
on the left bank of the Stolbovaya River in its lower reaches. Several
trees have been tilted under the impact of ice drift during the flood in
the spring of 2002.
Thus the following processes can be singled out as
likely rewonses to climate warmina:
I- ~
increase of the active layer thickness and intensification
of solifluction processes;
local replacement of solifluction processes by
landslide mass movement in the areas of active
river erosion;
frequent falling of trees with creeping root system.
This usually happens where the hydromorphic
clays having a viscous-plastic consistence reach a
thickness of about I .5 m or more
presence of trees with bent trunks and vertical tops;
improvement of drainage on top surfaces and
adjacent gentle slopes (water disappeared from
rupture-windows; it is often absent in fissuresruptures);
increase in the number of blocks in unstable
positions on the surface of kurums as well as of
size and number of reindeer moss patches;
exhaustion of ground rills under block cover of ku-
This research is carried out within the framework of RFBR
project 05-01 -64901.
REFERENCE
Gorshkov, S., Vandenberghe, J., Alexeev, B., Mochalova, 0. and
Tishkova, M. 2003. Climate, permafrost and landscapes of middle
yenisei region. Faculty of Geography of M.V.Lomonosov Moscow
State University, Moscow, (In Russian and in English).

8th International Conference on Permafrost Extended Abstracts an Current Research and Newly Available ~ n ~
New permafrost studies at Lake Hovsgol, north-central Mongolia
C. Goulden, *N. Sharkhuu, Ya. Jambaljav, **B. Etzelmiiller, E. S.F. Heggem, ***V. Romanovsky,
****R. Frauenfelder, A. Kaab, *****l. Ariuntsefseg, N. Saruul and S. Tumentsefseg
Mongolian Institute, Academy of Natural Sciences, Philadelphia, U.S.A.
*Institute of Geography, Mongolian Academy of Sciences, Ulaan Baatar, Mongolia
**Department of Physical Geography, University of Oslo, Oslo, Norway
***Geophysical Institute, University of Alaska, Fairbanks, U. S.A.
****Glaciology and Geomorphodynamics Group, Department Geography, University of Zurich, Zurich, Switzerland
*****HovsgoI GEF project, Geoecology institute, Mongolian Academy of Sciences, Ulaan Baatar, Mongolia
Lake Hovsgol in northern Mongolia occupies a north-south
aligned tectonic basin at the southern end of the Baikal
Rift System at an elevation of 1645 m, in a biogeographic
and ecological transition zone from taiga forest to steppe
and at the southern edge of the continuous permafrost
zone (Fig. 1). Hovsgol is a Mongolian Long Term Ecological
Research network site (Goulden et al. 2000). In
2002 a research program funded by a five-year grant from
the Global Environment Facility to the Geoecology Institute
of the Mongolian Academy of Sciences was initiated
to define the impacts of deforestation, nomadic pasture
use, and climate change on soil and permafrost conditions,
and plant and animal diversity. Analysis of long-term
temperature data from the Hatgal meteorological station at
the south end of the Lake indicates that annual temperatures
have warmed by about 1.5 "C since 1963.
Figure 1. Key map and study area of Lake Hovsgol GEF monitoring
and research valleys. The area lies on the eastern shore of the lake.
Six valleys along the northeastern shore of the Lake
have been selected for study, ranging from the heavily
grazed northern valleys, Turag (51" 17'46" N,
lOO"47'51"E) and Shagnuul, south of Hanh, Noyon and
Sevsuul with moderate grazing, and Dalbay and Borsog
(50°58'24"N, IOO"44'54"E) valleys in the south with little
N
or no grazing pressure (Fig. I). This gradient of pastoral
use allows us to define impacts of climate change in the
south, and its combined effects with pastoral use in the
north on biodiversity and ecosystems. The valleys have a
similar alignment, flowing from east to west, with northfacing
forested slopes covered by larch (Larix sibirica) and
south-facing mixed forest and steppe slopes and steppe
valley bottoms. Two meteorological stations monitor climate
conditions, one in Dalbay Valley, and the second at
Hatgal.
The goals of the permafrost studies are to define the distribution
and temperature of permafrost, the depth of active
layers in detail, and begin a monitoring program of soil
and permafrost temperatures in each of the valleys. In
June 2002, N. Sharkhuu drilled six boreholes to a depth of
six meters at different topographic locations in the Borsog
and Dalbay valleys. In late August to early September
2002, studies of the distribution of permafrost and active
layer depths were initiated, and temperature data loggers
were placed in two watersheds to add to information
gained from boreholes for monitoring soil and permafrost
temperatures in the study area (Fig. 1, Etzelmuller et al.
2001).
Sandy soils dominate on slopes but thick sequences of
lake sediment layers cover the valley bottoms. 1 D resistivity
soundings were carried out at four sites, and 2D resistivity
tomography at ten sites in the field. Resistivities were
generally low, calibration with borehole temperatures indicate
permafrost conditions at resistivities close to I kQm,
while ice-rich sediments in the valley bottoms showed resistivities
up to 10 kQm. The results have shown that
permafrost occurs on north-facing slopes and in valley
bottoms. Permafrost thickness in the valley bottoms is estimated
to be between IOand 15 m, while active layer
thickness was around 3-5 m on dry locations and down to
below 1 m in wet, vegetation-covered sites. South-facing
slopes have little or no permafrost. In valleys where intense
grazing occurs, soils are very dry, resulting in relatively
high resistivities even in the active layer. In these
45

Goulden, C. et al.: Permafrost studies, north-central Mongolia ...
/
a a d t h i c k a c a a
Figure 2. Preliminary permafrost profile of Borsog and Dalbay valleys, Lake Hovsgol, northern Mongolia (from N. Sharkhuu). For an overview of
recent permafrost changes in Mongolia, see Sharkhuu (2003).
areas, more borehole information is required; additional
boreholes will be drilled in early 2003 (Fig. 2).Two MRC
soil temperature measurement probes were installed at
the two meteorological stations, one on the north-facing
slope in the Dalbay River valley, and the other at the Hatgal
meteorological station. These temperature probes will
provide year-round hourly soil temperature data at eleven
different depths (approximately every IOcm), starting at
the ground surface and down to one meter. These data,
together with temperature data from the deeper boreholes,
will be used for calibration of a numerical thermal model
developed by the Permafrost Lab, Geophysical Institute,
UAF (Romanovsky et al. 1997, Romanovsky and Osterkamp
2001). The site-specific calibrated models will be
applied to the entire period of available meteorological
data from this region to reconstruct the permafrost temperature
dynamics within different landscape units.
Transects for the study and monitoring of plant taxa
and biomass growth, terrestrial insects, birds, mammals,
and stream flow, water chemistry and aquatic biology
have been set across each valley and longitudinally along
the streams. Detailed studies in Borsog Valley show that
increased soil moisture occurs on lower slopes associated
with permafrost and plant and insect taxa and biomass
vary with soil moisture. In Turag and Shagnuul with very
dry soil, grazing may have affected permafrost conditions,
plant biomass is low but plant taxa number is high apparently
due to an increase in the number of xerophytic plant
taxa. Experimental studies will begin in 2003 to study the
impact of plant cover on soil and permafrost temperature.
REFERENCES
Etzelmuller, B., Holzle, M., Soldjorg Flo Heggem, E., Isaksen, K.,
Mittaz, C., Vonder Muhll, D., mdegard, R. S., Haeberli, W. and
Sollid, J. L. 2001. Mapping and modeling the occurrence and distribution
of mountain permafrost. Norsk Geografisk Tidskrift, 55(4):
186-1 94.
Goulden, C. E., J. Tsogtbaatar, Chuluunkhuyag, W. C. Hession, D.
Tumurbaatar, Ch. Dugarjav, C. Cianfrani, P. Brusilovskiy, G.
Namkhaijantsen and R. Baatar. 2000. The Mongolian LTER:
Hovsgol National Park. Korean J. Ecol. 23:135-140.
Romanovsky, V.E., Osterkamp, T.E. and Duxbury, N. 1997. An
evaluation of three numerical models used in simulations of the
active layer and permafrost temperature regimes, Cold Regions
Science and Technology. 26 (3): 195-203.
Romanovsky, V. E. and Osterkamp, T. E. 2001. Permafrost: Changes
and Impacts, in: R. Paepe and V. Melnikov (eds.), Permafrost Response
on Economic Development, Environmental Security and
Natural Resources, Kluwer Academic Publishers: 297-315.
Sharkhuu, N. in press. Recent changes in the permafrost of Mongolia.
Proceedings, 8th International Conference on Permafrost. Zurich,
Switzerland.
46

8th ~ nt~r~~~i~nal Conference on Permafrost Extended Abstracts on Current Research and Newly Available l ~ f ~ ~ ~ a ~ l ~ n
Snow-free ground albedo derived from DAIS 7915 hyperspectral imagery
for energy balance modeling in high-alpine topography
S. Gruber, *D. Schlapfer and M. Hoelzle
Glaciology and Geomorphodynamics Group, Department of Geography] University of Zurich, Switzerland
*Remote Sensing laboratories, Department of Geography, University of Zurich, Switzerland
INTRODUCTION
In alpine topography, air temperature, solar radiation,
snow cover height and duration exhibit a strong lateral
variability. Statistical models (e.g. Gruber and Hoelzle
2001) are only partially able to approximate these processes
with a limited number of input variables such as
mean annual air temperature and simulated solar irradiation.
Physically-based distributed modeling of the vertical
energy balance can realistically simulate complex interactions
of several variables as demonstrated by Stocker-
Mittaz et al. (2002). Their model PERMEBAL simulates
the daily vertical energy balance, based on ground characteristics
and meteorological time series. The required
base data such as albedo, emissivity and roughness
length is needed for every grid cell and must therefore be
measured or accurately estimated. During snow-free conditions
in summer when energy input to the ground is
largest, the daily albedo mean is nearly constant. Remote
sensing is the only way to provide corresponding spatial
data fields. This study uses imaging spectrometry together
with geo-atmospheric correction techniques to measure
albedo in a complex topography. The test site is the
Corvatsch area, Upper Engadin, Switzerland, where the
model PERMEBAL was developed.
METHODS AND MEASUREMENTS
Albedo is defined as the ratio of outgoing over incoming
radiation and, for climatological purposes, it is usually
considered in the wavelength interval of about 300 to 2500
nm, outside of which solar irradiation is small. For periglacia1
surfaces without snow cover, albedo usually lies between
0.1 and 0.3. In the determination of albedo with remote
sensing, ground reflected radiance as well as
irradiance needs to be quantified. As a consequence, a
well-known and well-calibrated sensor is needed together
with accurate characterization of the atmosphere and solar
illumination geometry. Because of the high spatial variability
of surface characteristics and illumination, all auxiliary
and image data must be accurately referenced.
In a raw image, sun-exposed slopes will generally appear
brighter than other areas. Haze in a valley may
lighten-up the valley bottom. However, these differences
largely reflect the influence of atmospheric and illumination
conditions but not the surface property under investigation.
The geo-atmospheric correction needs to account
for this effect and convert an at-sensor radiance image to
reflectance. However, this reflectance (Le. hemisphericaldirectional
reflectance factor, HDRF) is still not suitable to
describe albedo because the reflectance of natural objects
usually exhibits a pronounced angular anisotropy. It is
mathematically described by the bi-directional reflectance
distribution function (BRDF). Multiangular observations of
a target and the use of a BRDF model is necessary to
quantitatively take into account angular anisotropy. Alternatively,
BRDF effects can be corrected using empirical
models. Both approaches are used in this study to derive
spectral albedos for each band. Broadband albedo can
then be calculated by spectral integration.
On 14 August 2002, the test area was covered by
three flight lines of the imaging spectrometer DAIS7915
under cloud-free conditions. The airborne DAIS7915 sensor
has 72 bands between 500 and 2500 nm and seven
additional infrared bands.
PROCESSING AND RESULTS
The system calibrated data has been geometrically corrected
using the software PARGE (Schlapfer & Richter,
2002). Aircraft position and attitude as well as a digital
elevation model are used to reconstruct the ground-sensor
geometry and precisely locate individual pixel footprints.
The final pixel size is 10 m and the standard errors in horizontal
accuracy were below 5 m. Atmospheric correction
has been performed for the image bands between 500
and 2500 nm wavelength using the software package AT-
COR4 (Richter and Schlapfer 2002). The effects of path
radiance, transmittance and direct and diffuse solar flux in
complex topography are accounted for.
In one experiment, the BRDF effect is empirically corrected
based on illumination geometry and using the algo-
47

ithm provided by ATCOR4. The correlation between solar
illumination intensity and the derived albedo serves as a
means to evaluate the removal of topographic effects as,
in theory, they should be virtually independent. Figure 1
demonstrates the success in removing the effects of illumination
and BRDF. The analysis is based upon 56,000
pixels of periglacial surfaces without snow. During this
empirical correction, however, the quantifiable physical
basis for the albedo product is lost. A comparison with
field measurements from one site gives good agreement,
280 pixels on the Murtel rock glacier had an albedo of
17.1 k 1.4% and the snow-free albedo derived from a
CM3 pyranometer (0.3 - 3 um) is 17% (Stocker-Mittaz et
al. 2002).
W
0
a,
P
0.6
0.5
0.4
E 0.3
m
E
0.2
0.1
0.0
0.0 0.2 0.4 0.6 0.8 1 .o
cosine of solar illumination angle
Figure 1. Dependence of albedo on solar illumination without topographic
correction (thin line), with topographic correction (crosses) and
with topographic correction and empirical BRDF compensation (thick
line).
In a second experiment (see Gruber et al. 2003 for
details) a BRDF model was fitted to the data and the effects
of angular anisotropy were explicitly corrected. The
parameterization of albedo by a BRDF model provided the
opportunity to investigate the influence that angular anisotropy
has on the net short-wave radiation in a complex terrain.
This was achieved by coupling the energy-balance
model PERMEBAL to a BRDF-based model of albedo.
Using meteorological data of the Corvatsch area, direct
and diffuse downwelling short-wave, radiation as well as
solar incidence angles were parameterized hourly for the
entire year 1999. Slope angles between 0 and 50" northand
south facing were simulated. The solar incidence angle
and the ratio of diffuse and direct radiation served as
input for the parameterization of albedo. The diurnal variation
of albedo measured by the climate station could successfully
be reproduced using this combination of remote
sensing data and PERMEBAL. Unfortunately, the accuracy
of the derived albedo is between 15 and 25%, only. It
was demonstrated that the same surface material has a
distinctly different albedo on slopes of different orientation
due to the dependence of albedo on solar incidence angles.
This difference is up to 25% and, therefore, an an-
gular parameterization of albedo should be considered in
complex terrain for accurate models of net short-wave radiation.
CONCLUSION AND OUTLOOK
The derivation of accurate albedo in complex terrain is still
subject to large errors, especially because of the difficulty
to quantify the BRDF for each pixel. The dependence of
albedo on solar incidence angle can lead to distinctly different
albedo values for the same surface in differing topographic
situations. This makes research into angular
anisotropy of albedo especially interesting for energybalance
modeling in complex topography. The variability
ranges of albedo as well as the achieved benefit to model
accuracy can now be assessed based on the data generated
in this project. The two BRDF corrections using an
empirical model and a BRDF model can be compared.
Furthermore, the data generated in this campaign will be
used to assess the error inherent in derivations of albedo
from operational orbiting satellites such as Landsat, SPOT
or ASTER.
ACKNOWLEDGEMENTS
This work has been supported by the Swiss National Science
Foundation project "Analysis and Spatial Modelling
of Permafrost Distribution in Cold-Mountain Areas by Integration
of Advanced Remote Sensing Technology". The
EU-funded project HySens gave additional support.
REFERENCES
Gruber, S. and Hoelzle, M. 2001. Statistical Modelling of mountain
permafrost distribution I local calibration and incorporation of remotely
sensed data. Permafrost and Periglacial Processes 12(1):
69-77.
Gruber, S., Schlaepfer, D. and Hoelzle, M. 2003. Imaging Spectrometry
in High-Alpine Topography: The Derivation of Accurate
Broadband Albedo. Proc. the 3rd EARSeL Workshop on Imaging
Spectroscopy, Oberpfaffenhofen, May 13-16 2003.
Richter, R. and Schlapfer, D. 2002. Geo-atmospheric Processing of
Airborne Imaging Spectrometry Data Part 2: atmosphericltopographic
correction. International Journal of Remote
Sensing 23: 2631-2649.
Schlapfer, D. and Richter, R. 2002. Geo-atmospheric Processing of
Airborne Imaging Spectrometry Data Part 1 : Parametric Orthorectification.
International Journal of Remote Sensing 23: 2609-2630.
Stocker-Mittaz, C., Hoelzle, M. and Haeberli, W. 2002. Modelling alpine
permafrost distribution based on energy-balance data: a first
step. Permafrost and Periglacial Processes 13(4): 271-282.
48

Permafrost conditions in the starting zone of the Kolka-Karmadon rocklice
slide of 20 September 2002 in North Osetia (Russian Caucasus)
W. Haeberh, C. Huggel, A. Kaab, *A. Polkvoj, "*/. Zotikov and **N. Osokin
Glaciology and Geomorphodynamics Group, Deparfment of Geography, University of Zurich, Switzerland
*Geological Survey, Environment Department, North Osetia, Vladikavkas
"Russian Academy of Sciences, Moscow
On the evening of 20 September 2002, a large rocklice
slide took place in the norh-northeast wall of the summit of
Dzhimarai-khokh, northern slope of the Kazbek massif,
Northern Osetia, Russian Caucasus. Large parts of the
debris covered Kolka Glacier at the foot of the slope
sheared off as a consequence of the main impact and
enlarged the initially detached mass by about 60 million
m3 of snowlfirnlice, debris, morainic material and water.
After sliding down at an average speed of about 200 kmlh
along a path of 18 km, the moving mass was strongly
compressed at the narrow entrance of the Genaldon
gorge (ca. 1300 m as.l.), ejecting a mudflow which continued
for another 15 km into the Giseldon valley. The
compressed avalanche deposit dammed local rivers and
led to the formation of several lakes. The largest of these
lakes inundated parts of the village of Saniba at the orographic
right valley side near Karmadon and constituted a
growing threat to the population of the downvalley town of
Gisel (Kaab and others 2003, Popovnin and others in
press).
The situation necessitated the immediate establishment
of an observationlalarm system at the dammed
lakes in order to protect downvalley settlements from potential
floods and avoid unnecessary evacuation or uncontrolled
reactions of inhabitants. At the same time, the
potential for an immediate repetition of similar or even
larger accidents from the mountain slope had to be assessed.
This involved interpretation of photographs collected
by the authorities during reconnaissance flights by
helicopter together with some best-guesses about thermal
aspects related to firn, ice and permafrost conditions in the
starting zone. The fact that two large slides from the same
slope but with considerably shorter runout distances had
occurred in 1902 - almost exactly 100 years ago - constituted
an important aspect of this assessment.
The avalanche starting zone as depicted in Figure I
was approximately 1 km wide and was located between
about 4300 m and 3500 m a.s.1. A roughly estimated total
of some 4 million m3 of steeply inclined metamorphic rock
layers were detached to a depth of about 40 m, entraining
a similar thicknesslvolume of snow, firn and glacier ice.
The dip of the bedrock layers, as visible in the detachment
zone after the event, was less than the inclination of the
now-exposed slope surface: outcropping of steeply inclined
bedrock layers in the lower part of the affected
slope (arrow in Fig. 1) could have constituted an important
geological influence on stability conditions. The primary
cause of the instability must have been within bedrock
rather than surface ice. As bedrock stability can be especially
low in warm or degrading permafrost (Davies and
others 2001), thermal conditions affecting ice and water
within rock fissures are likely to have exerted major and
detrimental influence.
Figure 1. Starting zone of the Kolka-Karmadon rocWice slide in the
north-northeast wall of the summit of Dzhimarai-khokh. Arrow points
to steeply inclined bedrock layers.
A weather station near Karmadon was run between
1962 - 1987, showing a mean annual air temperature
(MAAT) of 59°C. Assuming a warmer recent MAAT of
about 7°C at 1300 m a.s.l., and a MAAT-lapse rate of 0.6
f O.I"CllOOm, present-day mean annual air temperature
can be estimated at -6 f 2°C at the lower and -11 f 3°C
at the upper end of the detachment zone. The existence of
active rock glaciers in the region (Fig. 2) indeed confirms
that the lower boundary of discontinuous mountain per-
49

Haeberli, W. et al.: Permafrost conditions in the starting zone of the Kolka-Karmadon rocklice slide ...
mafrost is around 2500 m a.s.l., where MAAT is slightly
below 0°C. Direct minilogger measurements in the Alps
(Gruber and others 2003) indicate that mean annual rock
temperatures tend to be somewhat higher than air temperatures
extrapolated from weather stations. This effect,
however is to some degree compensated for by the fact
that the affected slope is oriented away from the sun and
hence will be colder than average. In the absence of any
direct measurements or numerical model results, bedrock
surface temperatures in the detachment zone may be estimated
at about -5 to -1 O T , indicating bedrock conditions
of cold permafrost after the event.
tioned by reconnaissance teams, documents that thermal
water surfaces in the critical zone.
The answer to the question about the potential for an
immediate repetition of a similar or even larger event was
based on the facts that (I) the elimination of hanging glaciers
with warm firn areas had reduced the load on the
slope and eliminated the main meltwater source, (2) the
exposed bedrock would now be subject to strong cooling
and deep freezing, (3) the sheared-off iceldebris volume
of Kolka glacier had made large mass increase at the foot
of the slope impossible and (4) the path of the 20 September
event with its heavy destruction was practically inaccessible
anyway. It was concluded that the threat from a
possible repetition at the same site and in the immediate
future of an even bigger accident could be considered
minimal but that instabilities in the remaining parts of the
slope could continue and should be observed accordingly.
The build-up over long time intervals of hanging glaciers
with complex thermal conditions, in combination with
thermal water in the region and the specific geological
setting (steeply inclined rock layers, volcaniclgeothermal
processes), could explain the "quasiperiodicity" of events
at a century time scale. Effects of potential future atmospheric
warming would also have to be taken into account
for the coming decades.
ACKNOWLEDGEMENT
Figure 2. Upper avalanche path of the Kolka-Karmadon rock/ice slide,
the summit of Dzhimarai-khokh and the detachment zone are in the
background. An active talus-derived rock glacier (arrow) is seen in the
lower right corner. Note steamldust () cloud at the foot of the slope
where Kolka glacier has been sheared off (star) and overriden tongue
of Maili glacier (circle). Kazbek volcano would be to the left.
During the years before the event, the slope had been
covered by a number of well-developed, spectacular
hanging glaciers. Such firnlice bodies are known to have a
complex distribution of englacial temperatures: while verticallimpermeable
cliffs, where ablation takes place through
ice breaking off, can be as cold as the surrounding bedrock,
permeable firn layers underneath the less inclined
upper surface with snow accumulation are strongly
warmed up by latent heat from percolating and refreezing
meltwater. In areas with MAAT exceeding -10 to -12"C,
firn is usually temperate and the icelbedrock interface behind
the cold cliff remains at phase equilibrium temperature,
a pattern, which introduces deep-seated thermal
anomalies within the underlying bedrock (Haeberli and
others 1997). It is, therefore, highly probable that the detachment
zone at Dzhimarai-khokh was in a complex condition
of relatively coldlthick permafrost combined with
warm if not unfrozen parts and meltwater flow in very
steeply inclined materials with most heterogenous permeability.
This situation favouring high and strongly variable
water pressures was further complicated by the fact that
hot springs are known to occur in this volcanic region of
the Kazbek massif: the steamldust () clouds visible on
most photographs, and the strong sulphur odour men-
The presented work resulted from collaboration with the
the Government of North Osetia and the Academy of Sciences,
Moscow, as part of a project directed and funded
by the Swiss Development Corporation - Humanitarian
Aid.
REFERENCES
Davies, M., Hamza, 0. and Harris, C. 2001. The effect of rise in mean
annual air temperature on the stability of rock slopes containing
ice-filled discontinuities. Permafrost and Periglacial Processes 12:
137-144.
Gruber, S., Peter, M., Hoelzle, M., Woodhatch, I. and Haeberli, W.
2003. Surface temperatures in steep Alpine rock faces - a strategy
for regional-scale measurements and modelling. Proceedings of
the 8th International Conference on Permafrost, Zurich, Switzerland.
Haeberli, W., Wegmann, M. and Vonder Muhll, D. 1997. Slope stability
problems related to glacier shrinkage and permafrost degradation
in the Alps. Eclogae geologicae Helvetiae 90: 407-414.
Kaab, A., Wessels, R., Haeberli, W., Huggel, C., Kargel. J.S. and
Khalsa, S.J.S. 2003. Rapid Aster imaging facilitates timely assessments
of glacier hazards and disasters. EOS 13/84: 117, 121.
Popovnin, V.V., Petrakov, D.A., Toutoubalina, O.V. and Tchernomorets,
S.S., 2003. Catastrophe glaciaire 2002 en Ossetie du
Nord. Unpublished french translation, original article to appear in
Earth Cryosphere VI1 (1).
50

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
The thermal regime of the coarse blocky active layer at the Murtel rock
glacier in the Swiss Alps
S. Hanson and M. Hoelzle
Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
Energy balance measurements on the Murtd rock glacier
in the Upper Engadine in the Swiss Alps have shown that
the sum of all measured components indicate a non-zero
energy budget (Mittaz et al. 2000). This pattern was seen
Haeberli (2000) stated that due to pronounced lateral
energy fluxes in inclined active layers consisting of coarse
blocks, pronounced differences can exist between mean
annual temperatures at the surface and at the permafrost
both in 1997 and in 1998 with positive deviations in the table, respectively. Harris and Pedersen (1998) suggest
winter months and negative deviations during the summer.
Already at this time it was suggested that the imbalance of
that the air moving through the voids between the large
blocks play a significant role in modifying the thermal rethe
energy exchange fluxes may be explained by un-
gime of the blocky surface material. They showed, like
measured advective energy fluxes within the active layer.
Furthermore, a de-coupling between the surface tem-
Humlum (1997), how the active layer protects the permafrost
core acting like a thermal filter when snow cover is
perature and the temperature within the active layer of the thin or absent and low wind speeds prevail. The summer
rock glacier was recognized which does support this inter- warming of the permafrost surface is prevented by the
pretation.
trapped cold air between the boulders in the active layer to
the benefit of the permafrost. Furthermore evaporative
cooling in summertime keeps the temperature in the active
layer low, A reverse of these processes caused by an
abrupt increase in air temperatures within the active layer
has also been reported. Humlum (1997) pointed out the
scenario with high wind speed and absence of snow cover
where forced ventilation caused by wind pumping disturbs
the pattern and an introduction of relatively warm air can
Figure 1. The Murtel rock glacier. The square indicates the position of
the measuring site. Photo: C. Rothenbuehler
The existence of permafrost in the Alps is strongly influenced
by incoming short-wave radiation and snow
cover thickness and durations. Prediction of present and
future permafrost distribution in the Swiss Alps is based
mainly on the prediction of the extent of these factors.
However, the above-mentioned result may indicate processes
related to permafrost conditions, especially in
coarse blocky material, that are not fully accounted for today
and could be important in coupling the atmosphere to
the ground when applying climate change scenarios.
have a deepening effect on the active layer. Latent heat
release caused by refreezing of meltwater during spring
can possibly also play an important role in heating up the
active layer.
The importance of each of these non-conductive processes
is highly dependent upon local site conditions and is
also highly variable. Many of the non-conductive processes
are seasonal and effective only during periods of
phase changes when the driven gradient near the ground
surface is relatively large. In the context of global warming,
a better understanding of the non-conductive proc-
esses in the active layer of alpine permafrost and how
these are related to seasonal alterations is required in order
to comprehend and anticipate the influence of climate
changes in the Alps.
The year 2002 featured two very different cold periods
with snow cover. The winter of 2001/2002 was very cold
and dry. The snow did not arrive before late January 2002
and the snow cover was thin and discontinuous. This led
to open voids between the boulders during the whole
winter, allowing free connection between the atmosphere
51

and the active layer. The winter of 2002/2003 started already
in October 2002 with heavy snowfall that covered
the blocky surface completely with several meters of snow
that lasted the year out. The thick snow cover sealed off
any exchange between the active layer and the atmosphere.
These two situations give an unique opportunity to
compare two extreme conditions on the snow cover’s influence
on the processes governing the energy fluxes in
the active layer.
REFERENCES
Haeberli, W. 2000. Modern research perspectives relating to permafrost
creep and rock glaciers: a discussion. Permafrost and
Periglacial Processes 11 : 290-293.
Harris, S. A. and Pedersen, D. E. 1998. Thermal regime beneath
coarse blocky material. Permafrost and Periglacial Processes 9:
107-120.
Humlum, O., 1997: Active layer thermal regime at three rock glaciers
in Greenland. Permafrost and periglacial processes, 8: 383-408.
Mittaz, C., Hoelzle, M. and Haeberli, W. 2001. First results and interpretation
of energy-flux measurements of Alpine permafrost. Annals
of Glaciology 31: 275-280.
Figure 2. The energy balance climate mast on the Murtbl rock glacier.
Photo: M. Hoelzle.
The aim of this project is to discuss the thermal regime
of an active layer at a high resolution in a coarse bouldery
environment in an alpine discontinuous permafrost setting.
A fully-equipped energy balance station has been logging
data every 30 minutes at the Murtel rock glacier during the
entire year of 2002. In the same period, seven thermistors
placed within the active layer of the rock glacier were
measuring temperatures at a 5-minute scale. Two of these
thermistors were measuring the air temperature between
the blocks. The remaining five thermistors were drilled into
the blocks. With the comparison of the meteorological observations
and the recorded temperature within the active
layer, a discussion is presented and data provided of the
processes that govern the state of the active layer. Indications
of non-conductive processes are analyzed in detail
with the temperature gradient within the active layer, and
the cumulative daily heat accumulation at the surface and
within the active layer itself.
52

8th International Conference on Permafrost Extended Abstracts an Current Research and Newly Available Information
Detection of permafrost structure by transient electromagnetic methoc 1 in
Mongolia
K. Harada, *K. Wada and **M. Fukuda
Miyagi Agricultural College, Sendai, Japan
*Mitsui Mineral Development Engineering Co. , 1 td. , Tokyo, Japan
**/nstitute of low Temperature Science, Hokkaido University, Sapporo, Japan
Field observations were carried out in order to examine
the applicability of the transient electromagnetic (TEM)
method to detection of permafrost structure in a discontinuous
permafrost area of Mongolia.
In Mongolia, frozen ground occupies nearly 63% of the
total territory of this country, and the thickness of permafrost
varies between 5 and 200 m (Choibalsan 1998). The
TEM survey was made in Nalaikh (47O45' N, 107°15' E),
sandstone. Volcanic ash and Quaternary deposit are
above it for approximately 300 m. The soil profile up to 6.8
m is clay and silt confirmed by the boring survey made in
November 1998.
The TEM survey was conducted near a borehole,
named #524. The measuring station interval was 10 m
and the length of the profile was 100 m, named St. 2900 -
St. 3000. St. 3000 was located I00 m from the borehole
-20 88 84 76 74 74 67 63 63 63' -20
~84 84
"r
% -60 ! I
I
-4 0
-60
,
I
-100 - -100
i 700
-120 I I I I I I -120
7
permafrost
base
I
i
- 1998/11/16
- 1999/6/3
2900 2920 2940 2960 2980 3000 -3 -2 -1 0 1 2 3
Distance (m)
Temperature ("C)
Figure 1, (a) Layered earth section near borehole #524 and (b) temperature profile in the borehole, Nalaikh. Values in figure (a)
show resistivity in ohm-m.
which is located in the suburbs of Ulaanbaatar, central #524. A central induction measurement configuration was
Mongolia, in November 1998. Around this area, perma- used in this site. The measurements were performed by
frost is discontinuously distributed and the mean annual using the transmitter loop of 60 x 60 m. The transient data
air temperature ranges between 4 and 0 'C (Choibalsan, were recorded using PROTEM 47 TEM system by Geon-
1998). Bedrock around this area is consisted of schist and ics Limited with a receiver coil having an effective area of
53

31.4 m At the borehole #524, the ground temperature to
the depth of 50 m was measured on 16 November 1998
and 3 June 1999. Based on the temperature profiles, the
permafrost base was located at the depth of 36.4 m (Fig.
1 b).
The resistivity structures near the borehole #524 are
shown in Figure I (a). Three- or four-layered structures
were obtained between St. 2900 and St. 2970. Rather low
resistivity values (40-100 ohm-m) were found, and the
variation of resistivity value was small. The first and third
layers had the lowest resistivity values, 42-62 ohm-m. The
resistivity value of the second layer was higher than that of
the other layers. At the stations between St. 2980 and St.
3000, a two-layered resistivity structure was estimated.
The resistivity values of the first and second layers were
almost constant, 63 and 55 ohm-m, respectively. The difference
of resistivity between the first and second layers
was very small. The lower boundarv of the first layer
ranged from 34 m to 45 m deep.
1
Harada, K. et a!.: Detection af permafrost structure by transient ...
periments (after Harada and Yoshikawa 1996, Harada and
Fukuda 2000, Harada et al. 2000). The experimental results
show that the resistivity difference of the fine-grained
soil is small due to the presence of unfrozen water, , especially
in case of low water content. Furthermore, the
measured ground temperature shows above -1 OC below
the depth of 10 m. As a resistivity of frozen soil is a function
of temperature, the resistivity difference becomes
small in this study site.
From the field observation, it is noted that the best-fit
model obtained by the TEM survey represents the permafrost
structure precisely at this site, even though the difference
in resistivity between frozen and unfrozen layers is
small.
REFERENCES
Choibalsan, N. 1998. Characteristics of permafrost and foundation
design in Mongolia. In Proceedings of 7th International Conference
on Permafrost: 157-160.
Harada, K. and Fukuda, M. 2000. Characteristics of the electrical resistivity
of frozen soils. Seppyo 62: 15-22 (in Japanese with English
abstract).
Harada, K., Wada, K. and Fukuda, M. 2000. Permafrost mapping by
transient electromagnetic method. Permafrost and Periglacial
Processes 11 : 71 -84.
Harada, K. and Yoshikawa, K. 1996. Permafrost age and thickness
near Adventfjorden, Spitsbergen. Polar Geography 20: 267-281.
-5 0 5
Temperature (OC)
Figure 2. Electrical resistivity of soil samples. q is volumetric
water content (after Harada and Yoshikawa 1996; Harada and
Fukuda 2000; Harada et al. 2000). Spitsbergen is the soil sample
obtained from Spitsbergen and CPC is from Caribou-Poker
Creeks, Alaska.
The surface layer represented permafrost as confirmed
by the boring survey. From the observation at the nearest
station of the borehole #524, St. 3000, the lower boundary
of the first layer was located at the depth of 34.2 m from
the best-fit model and the allowable depth ranged between
23 m and 74 m based on the equivalence analysis.
The lower boundary of the best-fit model, 34.2 m, was in
good agreement with the depth of permafrost base obtained
from the ground temperature profiles, 36.4 m.
Therefore, this boundary indicates the permafrost base.
The observational results show that the difference in resistivity
values between the frozen and unfrozen layers
was very small. This small difference may have been
caused by the low water content. Choibalsan (1998) reported
that Mongolian permafrost is generally characterized
by low to moderate ice contents. Figure 2 represents
the relationship between resistivity values of fine-grained
soils and temperature obtained from the laboratory ex-
54

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Application of a capacitively coupled resistivity system for mountain
permafrost studies and implications for the interpretation of resistivity
values
C. Hauck, and *C. Kneisel
Graduiertenkolleg Natural Disasters & lnsfifufe for Meteorology and Climate Research, University of Karlsruhe, Germany
“lnstitufe for Physical Geography, Universify
of Wurzburg, Germany
INTRODUCTION
particular frequency by the transmitting dipole. The resulting
voltage coupled to the receiver‘s dipole is measured. In or flat
Geophysical techniques have become more and more easily accessible terrain data collection is faster than when
popular to determine the physical properties of frozen ground, using conventional DC resistivity systems. Further details
as they are comparatively cheap and easy to apply on most are given in Timofeev et al. (1994).
kinds of permafrost terrain. Within the large variety of gee The DC resistivity tomography technique has been used
physical techniques, electric and electromagnetic methods
increasingly in permafrost studies in recent years. The details
are especially well suited for permafrost studies, as resistivity
of the method and the peculiarities of its application to mounincreases
with increasing ice content (or decreasing unfrozen
water content), enabling the determination of ice content and
tain permafrost are described in many recent publications
ground ice characteristics in principle. However, in practical (e.g. Marescot et al. 2003, Hauck et al. 2003) and will not be
applications a number of technical limitations often prohibit a repeated here.
reliable assessment of the ice content. Among these, the in- The test sites of this study are located in the Bever valley
fluence of the electrical contact of the electrodes (in DC resis- (eastern Swiss Alps, see Kneisel et al. 2000) and on a nontivity
surveys) and the influence of data inversion of the frozen grass field near St. Moritz, which was used as refermeasured
apparent resistivities are probably the most impor- ence area. DC resistivity measurements were conducted
tant. Data inversion includes the problem of equivalence, with a SYSCAL Junior Switch system with Wenner and Diwhich
may be addressed by giving ranges for the resistivity pole-Dipole configurations. The OhmMapper surveys were
values of the various layers (e.g. Kneisel 1998), and the in- conducted in Dipole-Dipole configuration using the same array
fluence of inversion parameters, which may change resistiv- geometries as for the SYSCAL surveys.
ity values over up to one order of magnitude (Hauck et al.
2003).
In this contribution we would like to address the discrep-
RESULTS
ancy of resistivity results obtained through different measurement
techniques by comparing data from several resistivity
Whereas very similar results were obtained with both methsurveys
along the same profile in the eastern Swiss Alps.
ods on the flat grass field of the non-frozen reference area (not
Hereby, a capacitively coupled resistivity system (Ohmshown
here), difficulties were encountered on the steep and
Mapper, Timofeev et al. 1994), which has been successfully
heterogeneous terrain in the Bever valley. The surface is
applied on Arctic permafrost, is applied on more heterogenevegetated
by small trees and bushes and covered partly by
ous mountain permafrost terrain. The applicability of this new
medium-sized boulders, which makes towing the OhmMapinstrument
was evaluated in comparison to a standard multiper
system a difficult task, as the antenna dipoles tend to get
channel resistivity system.
stuck between boulders and bushes. In addition, downslope
METHODS AND DATA ACQUISITION
The OhmMapper (Geometrics) is a capacitively coupled re
sistivity meter that measures the electrical properties of the
subsurface without the galvanic coupling of ground electrodes
used in traditional resistivity surveys. A simple coaxial-cable
array with transmitter and receiver sections is pulled along
the ground. An alternating current is induced in the earth at a
towing is only manageable if a second person is holding the
end of the line in order to prevent uncontrolled tumbling OT
misalignment of the dipole antennas (which is also the case
on snow-covered slopes).
Figure 1 shows the survey results for both methods in
comparison to the DC resistivity survey conducted in 1998
(Kneisel et al. 2000). All results are shown as a horizontal
variation of electrical resistivity along the survey line. In principle,
the 2002 results of the OhmMapper and the DC resistivity
survey (SYSCAL) with Dipole-Dipole configuration
55

Hauck, C. et al.: Application of a capacitivety cuupled rcsistivily system for rnountaine
shouldbethesame,asmeasurementsweretakenon the
same day, at the same location and with the same survey
geometry. Even though the general resistivity distribution is
similar, the OhmMapper results (Fig. la) have much larger
noise content than the SYSCAL data (Fig. I b). This is mostly
due to difficulties in keeping the correct distance between receiver
and transmitter, as well as the correct orientation of the
dipole antennas due to the steep and uneven terrain. Besides
and possibly more important, the absolute values of apparent
resistivities are systematically lower for the OhmMapper (1
kam) than for the SYSCAL (4 kam). In contrast, it is seen
that the differences between the SYSCAL Dipole-Dipole and
Wenner results (Figs. 1 b and IC) and the Wenner results ob
tained in different years (1998 and 2002, Fig. IC) are less
variable than between OhmMapper and SYSCAL (Figs. la
and 1 b) obtained at the same time and at the same locations. Figure 2. Comparison of measured apparent resistivity values obtained
with SYSCAL and OhmMapper. Values are given in
loglo(Wm).
Figure 1. Comparison of apparent resistivity variation along the
survey line in the Bever valley for (a) OhmMapper, (b) SYSCAL Dipole-Dipole
array for 2 different spacings and (c) Wenner array
from 2 different years.
The discrepancies between SYSCAL and OhmMapper
are larger the higher the resistivity values. Plotting the apparent
resistivity values obtained with the OhmMapper system
against the corresponding apparent resistivity values of the
SYSCAL system, systematic differences are seen for resistivity
values above 500 m with smaller values for the
OhmMapper system (Fig. 2). This systematic discrepancy is
more pronounced the larger the resistivities. For very high
resistivities the OhmMapper values reach less than for the
values obtained with the SYSCAL system.
The possible reason is quite likely found in the dflerent
electrical coupling methods of both techniques. The high contact
resistances typically found in mountain permafrost terrain
result in larger apparent resistivities than obtained with capacitively
coupled instruments (OhmMapper) or electromagnetic
induction systems, where no coupling is needed at all.
In most mountain permafrost applications only relative resistivity
variations are needed, as the aim is to detect lateral w
vertical resistivity anomalies indicating the presence of frozen
material. However, when comparing data from dfferent IOMtions,
care has to be taken as absolute resistivities may be
overestimated due to resistive surface characteristics at specific
locations.
REFERENCES
Hauck, C., Vonder Muhll, D. and Maurer, H. 2003. Using DC resistivity
tomography to detect and characterize mountain permafrost.
Geophysical Prospecting 51 (in press).
Kneisel, C. 1998. Occurrence of surface ice and ground
icelpermafrost in recently deglaciated glacier forefields,
St.Moritz area, eastern Swiss Alps. Proc. of the 7th Int. Conf. on
Permafrost, Yellowknife, Canada: 575-581.
Kneisel, C., Hauck, C. and Vonder Muhll, D. 2000. Permafrost below
the timberline confirmed and characterized by geo-electric resistivity
measurements, Bever Valley, eastern Swiss Alps. Permafrost
and Periglacial Processes 11 : 295-304.
Marescot, L., Loke, M.H., Chapellier, D., Delaloye, R., Lambiel, C.
and Reynard, E. 2003. Assessing reliability of 2D resistivity imaging
in permafrost and rock glacier studies using the depth of
investigation index method. Near Surface Geophysics l(2): 57-
67.
Timofeev, V.M., Rogozinsky, A.W., Hunter, J.A. and Douma, M. 1994.
A new ground resistivity method for engineering and environmental
geophysics. Proc.of the SAGEEP, Annual Meeting: 701-
715.
56

8th International Conference an Permafrost Extended Abstracts an Current Research and Newly Available l ~ f ~ ~ ~
Combining and interpreting geoelectric and seismic tomographies in
permafrost studies using fuzzy logic
C. Hauck and *U. Wagner
Graduiertenkolleg Natural Disasters & lnsfitute for Meteorology and Climate Research, University of Karlsruhe, Germany
*Graduiertenkolleg Natural Disasters & Institute for Program Structures and Data Organization, University of Karlsruhe, Germany
INTRODUCTION
For many problems in periglacial and glaciological studies,
but also for other environmental, engineering and archaeological
problems, geophysical investigations are the
main source of information concerning the structure and
characteristics of the subsurface. In particular, the ability
to gather 2-dimensional information about the subsurface
material is considered the main advantage of geophysical
techniques as opposed to the single-point information
through boreholes. However, the indirect nature of geophysical
surveys, where material properties like water
content, pore size or temperature have to be inferred from
the measured physical variables such as electrical resistivity,
can be considered a major drawback for the application
of geophysical methods in applied geosciences.
Nevertheless, the uncertainty in the interpretation of geophysical
data sets is seldom explicitly treated in geophysical
applications.
In order to address this uncertainty, a model approach
using fuzzy logic is presented in this contribution, which is
based on expert knowledge about a particular problem.
The problem is taken from the area of ground-ice detection,
where a large variety of geophysical methods has
been applied in recent years (e.g. Vonder Muhll et al.
2001, Hauck and Vonder Muhll 2003). Generally, more
than one method is used in order to get a multivalued data
set and to facilitate data interpretation (e.g., DC resistivity
and refraction seismics). However, due to the nonuniqueness
of the inversion results and the range of possible
values for most earth materials, the interpretation of
the data set can be very ambiguous.
FUZZY MODEL
Fuzzy logic is an extension of a multivalued logic, where
classes of objects have unsharp boundaries and membership
is a matter of degree. It is a convenient way to map
an input space to an output space, especially if a quantitative
output is required from imprecise input variables. In
this case the input space is comprised by the results from
the geophysical surveys (e.g., electrical resistivities or
seismic P-wave velocities) and the output space is the
“degree of ice content” (but not the ice content itself), that
is the possibility of ground-ice occurrence.
Figure 1, Decision surface for the fuzzy inference system for ground
ice detection. Each value of seismic velocity (in kmls) and the logarithm
of electrical resistivity (in loglo(Wm)) is associated with a degree
of possibility for ground ice (z-axis).
The fuzzy inference system used in this study is of the
so-called Mamdani-type (Mamdani and Assilian 1975) and
is based on nine rules, all linking low, medium and high
resistivity and velocity values to a corresponding output
(low, medium and high ice content). A practical view of the
rule system is shown as decision surface (Fig. I), where
each pair of resistivity and velocity data point is associated
with a possibility of ground-ice occurrence. High resistivity
values and medium seismic velocities around 3500 mls
(velocity for pure ice) are associated with a high possibility
of ice occurrence, as can be seen from Fig. 1. Due to the
exponential increase of resistivity with decreasing and
negative temperature, the logarithm of resistivity is used in
the model.
57

RESULTS
Hauck, C. et ai.: C ~ ~ b and ~ interpreting n i ~ ~ geoelectric and seismic tomographies ...
The decision surface and the corresponding fuzzy model
are tested using a synthetic data set of resistivity and
seismic data pairs (Fig. 2). Both data sets are comprised
of 4 regions with high, medium and low resistivity and velocity
values, respectively, where only the upper right hand
corner represents values consistent with the material
properties of ice. Accordingly, the resulting model output
shows a region of high possibility for ice occurrences in
this part, whereas the possibility is low for all other regions.
This outcome would have been difficult to predict by
“manual” inspection of the resistivity and seismic distribution
alone, as the two patterns are notably different.
Hauck, C. and Vonder Muhll, D. 2003. Evaluation of geophysical
techniques for application in mountain permafrost studies. Zeitschrift
fur Geomorphologie, Suppl., special issue: geophysics in
geomorphology (in press).
Kneisel, C., Hauck, C. and Vonder Muhll, D. 2000. Permafrost below
the timberline confirmed and characterized by geo-electric resistivity
measurements, Bever Valley, eastern Swiss Alps. Permafrost
and Periglacial Processes 11: 295-304.
Mamdani, E.H. and Assilian, S. 1975. An experiment in linguistic
synthesis with a fuzzy controller. International Journal of Man-
Machine Studies, 7 (I): 1-13.
Vonder Muhll, D., Hauck, C., Gubler, H., McDonald, R. and Russill, N.
2001. New geophysical methods of investigating the nature and
distribution of mountain permafrost. Permafrost and Periglacial
Processes, 12 (1): 27-38.
4.8
I
4ti
4.4
4.2
4
38
3.6
34
Figure 2. Resistivity (left, in loglo(Qm)) and seismic velocity (center, in
kmls) of the synthetic data set, and furzy model output (right, possibility
of ground ice).
As can be seen from Figure 2, the fuzzy model can be
applied to detect areas with high possibility for ground-ice
occurrence from joint resistivity-velocity data sets. In an
initial evaluation study, the model was applied to several
field data sets from anomalously cold low altitude sites in
the midlands of Central Europe and the Alps, where permafrost
is sporadic or absent in general. However, ground
ice may be present in isolated patches with specific microclimatic
conditions (Kneisel et al. 2000, Gude et al. 2003).
First results show anomalies with an enhanced possibility
for ground-ice occurrence in comparison to the neighbouring
areas. Even though neither the probability nor the
actual ice content can be determined by the model, the
possibility for ground ice can be quantitatively compared
between different the field sites.
As the fuzzy inference system and the governing decision
rules can easily be adapted to different types of input
data, the applicability of the fuzzy logic approach may be
wide.
REFERENCES
Gude, M., Dietrich, S., Mausbacher, R., Hauck, C., Molenda, R., Ruricka,
V. and Zacharda, M. 2003. Permafrost conditions in nonalpine
scree slopes in Central Europe. 8th Int. Conf. On Permafrost,
Zurich, 2003, in press.
58

8th International Conference on Permafrost Extended Abstracts an Current Research and Newly Available Information
The spatial extent and characteristics of block fields in Alpine areas:
evaluation of aerial photography, LIDAR and SAR as data sources.
S. Heiner, S. Gruber and *E. Meier
Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
*Remote Sensing Laboratories, Department of Geography, University of Zurich, Switzerland
INTRODUCTION
The coarse surface layer in block fields operates as an insulating
layer between the surface climate and the ground
below. Often it acts as a thermal filter, insulating against
the effect of surface warming but readily transmitting the
effect of surface cooling (Harris and Pedersen 1998).
Therefore, knowledge about the extent of coarse surfaces
is important for the successful spatial application and further
development of energy balance models (Stocker-
Mittaz et al. 2002) or statistical models (e.g. Cruber and
Hoelzle 2001).
Block fields can be mapped relatively accurately based
on the manual interpretation of aerial photography because
of the textural information and terrain position. This
is a labor intensive process depending on the skills of the
operator. This study uses high resolution aerial photography
and a GIS (Geographic Information System) to
automatically map the spatial extent and characteristics of
block fields. The test site is the Corvatsch area in the
south-eastern part of the Swiss Alps.
METHOD AND DATA
In aerial photographs, block fields are characterized by the
frequency of change between dark and bright areas
caused by the shadows of big blocks. Other areas are
typically much more homogeneous. This effect can be
used to separate blocky areas from areas with fewer
blocks by using statistical methods.
On 17 September 2002, aerial photographs of the test
area have been taken by the Federal Office of Topography
of Switzerland. The ground resolution of these pictures
is about 6 cm.
Two more methods for the remote sensing of surface
roughness appear feasible and are currently under investigation:
(I) a SAR (Synthetic Aperture Radar) image of
JERS (Japanese Earth Resources Satellite) is currently
processed and (2) a flight with a LIDAR (Light Detection
And Ranging, laser altimetry) sensor on board will take
place in the test area this summer. LIDAR data from a
different campaign is already available for testing the
methods.
The correctness of the interpretation and validation of
the extracted coarsenesses is provided by transects obtained
from terrestrial surveying.
PROCESSING AND RESULTS
The aerial photographs are orthorectified and imported
into the GIs. Focal functions are applied on this grid.
These are functions which calculate a new value for a cell
using the value of this cell and its adjacent cells. Three
new grids are calculated: one contains the maximum of
two adjacent cells, and the second the mean of two adjacent
cells. Then the difference between these two grids is
calculated and the sum of ten adjacent cells is written into
the central cell. As a result, areas with a high frequency of
change between dark and bright cells (block fields) will
show high values, whereas more homogeneous areas will
be attributed lower values. Now the grid will be reclassified
into the different classes of surface roughness. The last
step is the resampling of the pixels to the desired resolution
for permafrost modeling.
First results are satisfactory as shown in Figure I. The
block fields could be delineated and the method clearly
shows the potential to distinguish between several classes
of surface roughness. Problems remain in shadow areas.
CONCLUSION AND OUTLOOK
The result presented above clearly shows the potential of
the method for the automatic mapping of block fields using
standard GIS software packages. Problems remain to be
solved in the shadow areas. The availability of this data
will improve permafrost modeling.
The use of JERS SAR and LIDAR data will provide an
evaluation using two more data sources. SAR will be inexpensive
and provide large area coverage but only have
a low spatial resolution and only be able to distinguish a
limited number of roughness classes in alpine terrain. LI-
59

DAR will permit a high spatial resolution and accurate
characterization and quantification of surface structure but
requires significantly more cost and effort.
First attempts to use wavelets in the detection of surface
roughness have shown interesting perspectives and
applications of this method to aerial photographs and LI-
DAR data are underway.
Heiner, S. et ai.: The spatial extent and characteristics of block fields,,,
Figure 1. The picture shows the area around the rockglacier Murtel.
Four classes of surface roughness are determined: highest values are
found on the coarse surface of the rockglacier, whereas the homogenous
surface of the skislope shows lowest values.
REFERENCES
Gruber, S. and Hoelzle, M. 2001. Statistical modelling of mountain
permafrost distribution - local calibration and incorporation of remotely
sensed data. Permafrost and Periglacial Processes 12 (1):
69-77.
Harris, S. and Pedersen, D.E. 1998. Thermal regimes beneath coarse
blocky materials. Permafrost and Periglacial Processes 9: 107-
120.
Stocker-Mittaz, C., Hoelzle, M. and Haeberli, W. 2002. Modelling alpine
permafrost distribution based on energy-balance data: a first
step. Permafrost and Periglacial Processes 13(4): 271-282.
60

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available l ~ f ~ r ~ ~ ~
Thermal regime of coarse debris layers in the Ritigraben catchment, Matter
Valley, Swiss Alps
T- Herz, L. King and *H. Gubler
lnsfifut fur Geographie der JLU, Giessen, Germany
*ALPuG, Davos, Switzerland
INTRODUCTION
The objective of the current study is to investigate the
microclimate within coarse debris layers in high mountain
environments. To quantify the heat exchange between the
local atmosphere and the near-surface ground, special
attention is focused on thermal conditions. Further-more,
their influence on the distribution pattern of discontinuous
mountain permafrost is of specific interest. Based on the
research concept and site description given in Herz et al.
(in press), this poster presents and discusses the initial
data gathered from the measurement equipment
described below.
RESEARCH AREA
The data was recorded in close proximity to the meteorological
station erected in the Ritigraben block slope
(46"l I 'N, 7'51 'E, 2615 m a.s.1.). Block sizes in this area
range from 0.5 up to several cubic meters. For comparison
purposes, the finer-grained substrate types known
as "ski run" and "soil beneath alpine grassland" were also
incorporated in the measurement programme.
MEASURING SETUP
Local climatic conditions are recorded at a meteorological
station. The ground thermal regime is registered in a 30 m
deep borehole (cf. Hen et al. in press).
To measure rock and air temperatures within the block
layer, two types of temperature sensors were constructed.
In both cases the sensor element consists of a high-precision
platinum thin film thermo-meter Pt 1000 113DINB
(LxWxH = 10~2~0.25 mm). The sensors were connected
to a 5-channel-mini-logger (Type Wickenhaeuser
TL-LOG5) by a half-bridge circuit. The technical components
of the logger and the use of a high-precision reference
resistor in the circuit guarantee excellent long-term
stability of the temperature measurements with a theoretical
resolution of 0.00012"C. A four-point calibration
was done in an ice-water bath for 0°C or a circulated
water bath controlled by an analogue calibration
thermometer (resolution 0.01 "C) for IO,20 and 30°C,
respectively. The calibration yielded outstanding results,
so that the accuracy of temperature measurements can be
indicated as within at least 0.02"C.
Rock and soil temperature sensor elements were
sealed in closed high-grade steel tubes with a wall
thickness of 0.2 mm. Air temperature sensor elements
were fixed in open steel tubes in such a way, so that the
tip of the Pt I000 sticks out of the tube and is in direct
contact with the surrounding air.
Block layer rock and air temperatures are measured in
a profile of four consecutive measurement points. The
profile follows the drainage way of the local micro relief.
Each measurement point consists of two loggers with five
rock and air temperature sensors installed at different
depths. Four additional measurement points form two
cross-sections to the main profile to also include other
micro topographical positions in the block slope.
Surface and soil temperatures are also recorded in the
substrate types, "ski run" and "soil".
INITIAL RESULTS
Hourly meteorological and daily borehole temperature
data have been available since April 2002. Microclimatological
data exist so far for the nine-week period
from August 15th to October 18th 2002. Measurement
intervals were set to 5 minutes for air temperature and 15
minutes for rock and soil temperature sensors during this
period. The borehole temperature data presented in
Figure I display a marked asymmetry. The heat input
during summer penetrates down to only 3.5 m in spite of
extreme temperature gradients of up to 10"Clm. A sharp
bend in the maximum curve occurs at that depth. The
maximum temperature remains very close to 0°C during
the whole summer between 3.5 and 5 m depth.
Penetration of surface cooling during winter reaches down
to 14 m depth, tantamount to the depth of zero annual
amplitude. These differences in ground thermal properties
61

etween "summer" and "winter" conditions can only be
explained by phase change processes at a permafrost
table situated between 3.5 and 5 meters depth during
summer 2002.
Temperature [$I
-10 -5 0 5 10 15 20 25 30
0
Herr, T. et 81.: Thermal regime of coarse debris.. .
response on a surface temperature decrease, which is not
the case for the soil.
. ". . - . . . . " . .. . . . ." .
"-2 cm (block layer)
. . .". ".. ..
. -. . . " .
-- 140 cm @lock layer)
"-150 cm (soil)
2
4
6
8
10
12
E 14
5
16
[Zv&num temperature ~
I ' ,
Maximum temperature1
"Mean temperature
-
Figure 2. Comparison of block layer and soil temperatures
from September 1st to October 18th, 2002.
18
20
22
24
26
28
30
Figure 1. Ground temperature envelope, Ritigraben block slope, 261 5
m a.s.1. (April 2002 -January 2003).
The resulting thermal offset in ground temperatures
according to Goodrich (1978) is indicated by a mean
temperature of -0.84"C at 3.5 m depth, which is 15°C
colder than the mean temperature at 0-1 m depth. The
permafrost body can be characterized as "warm" with
mean temperatures of -0.37"C at depth of ZAA and
-0.61"C at 30 m depth. It is remarkable, that absolute
minimum temperatures were measured at 1.5 m depth,
representing the bottom side of the uppermost block of the
surface layer. The bottom side is obviously influenced by
advective air circulation between the block layer void
system and the near-ground atmosphere also under
winter conditions, while its surface is protected from heat
loss by the formation of a snow cover.
Figure 2 compares rock temperatures in the block layer
with temperatures measured in an adjacent soil profile.
The different behaviour of the two deeper curves is of
special interest in this context. Rock temperatures at 140
cm depth still show a daily course, which is not detectable
at a comparable depth in the soil. Altogether the temperature
level in the block layer is lower than in the soil in spite
of higher surface temperature values. Rock temperatures
at 140 cm depth clearly correlate with the daily minima of
the surface temperature course and show an immediate
CONCLUSIONS AND OUTLOOK
The data presented indicate that non-conductive heat
transfer processes play an important role in the thermal
regime of block layers. Advection of water and air in the
void system between blocks is believed to be the crucial
phenomenon, which is investigated in this study.
Furthermore, in October 2002 five additional boreholes
were drilled in the lower part of the Ritigraben block slope
to investigate the rheological conditions in the catchment
of the torrent.
REFERENCES
Goodrich, L.W. 1978. Some results of a numerical study of ground
thermal regimes. In 3rd International Conference on permafrost,
Proceedings, Edmonton, 10-13 July: 29-34.
Herz, T., King, L. and Gubler, H. in press. Microclimate within coarse
debris of talus slopes in the alpine periglacial belt and its effects
on permafrost. In 8th International Conference on Permafrost,
Proceedings, Zurich, 21-25 July 2003.
62

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Topographical properties of active and inactive patterned grounds in
Finnish Lapland
J. wort and *M. Seppala
Department of Geography, University of Helsinki, Finland
*Physical Geography Laboratory, University of Helsinki, Finland
Periglacial patterned grounds are quite widespread features in
Northern Finnish Lapland, especially above timberline in the
zone of discontinuous permafrost. The most common types
are sorted and non-sorted nets. Activity of these landforms
varies from one area to another and between different types of
patterned grounds. Active forms seem to be mainly located in
the valleys, Respectively, inactive patterned grounds are
more common on the slopes and on top of the fells (Fig. I). In
general, climate is a crucial environmental determinant of frost
activity. However, when we study regions at large scale the
importance of local factors such as soil, hydrology, topography,
vegetation etc., becomes apparent.
The objectiveis to present some results of a research
where active and inactive patterned grounds were studied
using Geographical Information System (GIS) and statistical
methods. The main aim was to compare topographical pmperties
of active and inactive regions to see if there are any
differences between these areas. Topographical parameters
calculated from a digital elevation model (DEM) were used
as explanatory variables in univariate analysis. The DEM
with 20 m grid size was created from contours of topographic
maps using Arcllnfo 8.1. SPSS 9.0 software package was
usedinstatisticalanalysis.
Figure 1. Distribution of active and inactive patterned grounds in a part (4.8 x 10 km) of the study area (10 x 20 km) at Paistunturit fell region
in Finnish Lapland (Contours 0 National Land Survey of Finland, Licence number 491MYY103).
63

Table 1. Results of an univariate analysis where active and inactive patterned ground squares were compared with topographical variables
calculated from DEM. Total number of the grid squares in the analysis is 745. Statistical significance is tested with the Mann-Whitney U-test.
Topographical Patterned ground Statistical
variables
(PI
Active (n 421) Inactive (n = 324)
mean f std
mean f std
Mean altitude (m a.s.1.) 392 f 40 454 f 49

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available ~~~~r~~~~~~
Influence of human activities and climatic change on permafrost at
construction sites in Zermatt, Swiss Alps
R. Hot 1. King, T. Hen and *S. Gruber
Institute for Geography, Justus Liebig University, Giessen, Germany
*Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
GEOGRAPHICAL BACKGROUND
PERMAFROST CONDITION
AND
(c) other related structures such as masts, tunnels, elevators,
shelters for vehicles, workshops, etc.
(d) subsurface water pipes (for drinking water, artificial
snowing of ski-runs), sewage, communication and electricity
Zermatt is a tourist resort in the Alps with a total of 14,000
hotel beds and more than one million oficial overnight stays lines.
per year. The necessary infrastructure consists of installations Engineering geologists as well as those responsible perthat
often reach into permafrost areas, from the sporadic zone
at about 2600 m a.s.1. up to the continuous permafrost zone
sons for this infrasttucture have become increasingly interested
in the distribution and the characteristics of permafrost in
above 3400 m a.s.1. The unglaciated permafrost area has a the Zermatt area, as there have been problems due to perma
large vertical extension due to the surrounding high mountain frost degradation.
ranges that reach above 4000 m a.s.l., resulting
a dry and The poster gives an inventory of the existing structures on
sunny climate and a very high glacier equilibrium line.
The infrastructure erected in the permafrost areas consists
of
probable and proven permafrost sites and describes some
problems encountered during the last 25 years, with original
new temperature data.
(a) hotels, restaurants and mountain huts,
(b) station buildings of railways, funiculars and ski lifts,
0 50m
Figure 1. Map of Kleinmatterhorn mountain peak with tourist facilities (IO meter contour lines).
x temperature logger
65

CONSTRUCTION SITES
Kulrnhotel and Gornergrat Funicular
Gornergrat can be reached by a 9.340 km long railway line
constructed between 1896 and 1898. The uppermost 500
meters were added in the year I909 and the Kulmhotel Gornergrat
was opened in 1910. The railway track and the hotel
were mainly used in summer until the early Fifties. Winter
skiing became more and more attractive in the Sixties when
the hotel started to open also in winter. In the year 1985 two
new astronomic observatories were added to the northern
and southern towers and this caused a substantial additional
load on the subsurface. Moreover, apartments for scientists
were heated in the basement.
Soon after this, the northern tower that was erected on
frozen morainic material started to settle due to permafrost
degradation. The subsidence between tower and hotel was
controlled with geodetic measurements and strain gauges and
concrete were injected into the foundation in order to stop the
differential settlement. Today, ground temperature measurements
show that the ground below the tower is probably not
frozen for a depth of about 10 meters (the active layer at un-
disturbed sites is less than 2.5 meters). This must be attrib
uted to the heating of the basement for almost 20 years. Actually,
subsidence seems to have stopped.
n of, R, et ai.: Human activities, climatic change, permafrost, Zermatt ...
In view of the considerable temperature rise in the bedrock,
temperatures havebeenmonitoredsince1998at ten
different places in order to takecountermeasures if necessary.
2.3 Further construction sites on permafrost
There are many more facilities erected on the permafrost (Table
l). Climatic change may further contribute to a rise in the
lower permafrost limit. It should be recalled that an actual
MAAT of about -1°C indicates patchy permafrost occurrences,
and at less than -5T, continuous permafrost may be
expected.
Table 1. Installations on permafrost in the
tudes and expected MAAT.
Zermatt area with alti-
Name
Altitude (m a.s.1.) MAAT (“C)
Kleinmatterhorn
(station) 3820 -7,2
(elevator summit) 3883 -7,6
Testa Grigia (station) 3479 -53
Stockhorn (station) 3407 -4,9
Hohtalli (station) 3286 -4,2
Matterhorn Refuge 3260 -4,O
Rote Nase (station) 3250 -4,O
Kulmhotel Gornergart 31 35 -3,3
Rothorn (station) 31 03 -3,l
Gandegg Refuge 3029 -2,7
Trockener Steg (station) 2939 -2,2
Gornergrat rack railroad (top) 3090 -3,O
CONCLUSIONS
Figure 2. Kleinmatterhorn mountain top with the location of temperature
loggers. The position of the cross sections A, B and C are
shown in Figure 1.
Degradationofpermafrostdue to climatic change may be
come a serious threat to installations at high mountain tourist
resorts. This is especially dangerous when the facilities are
not properly maintained. Therefore, relevant information must
be improved to the managers of these installations and to the
local authorities by permafrost scientists working in these areas.
Kleinmatterhorn Funicular
This funicular travels up to 3820 m a.s.1. and were constructed
in 1981, It arrives at a tunnel cut in the northern wall
of the mountain top (Fig. 1). At its southern exit the ski tun
starts down to Zermatt. Bedrock temperatures as low as
-12°C were reported during the construction (Keusen and
Haeberli 1983). The same temperature was measured in the
near-surface rock material of the northem rockwall below the
mountain top in a diploma study. The MAAT measured here
was -8°C in the years 1998199.
Today, bedrock temperatures have risen to -3 to -2°C at
several localities (cf. Fig. 2). This is due to heating and to the
heat brought into the tunnel by the more than 490,000 visitors
per year. Additional heat is created by almost 70,000 elevator
movements per year, transporting tourists to the mountain top.
In summer 1997, meltwater created problems when re
freezing in the elevator shaft (King et al. 1998).
REFERENCES
Keusen H.R. and W. Haeberli, W. 1983. Site investigation and
foundation design aspects of cable car construction in Alpine
permafrost at the “Chli Matterhorn”, Wallis, Swiss Alps. In 4Ih International
Conference on Permafrost, Proceedings, Fairbanks,
17-22 JuIv: 601-605.

8th International Conference an Permafrost Extended Abstracts an Current Research and Newly Available ~ n f ~ r ~ ~ ~ i o
Spatial conditions of frozen ground and soil moisture at the southern
boundary of discontinuous permafrost in Mongolia
M. Ishikawa, Y, Zhang, T. Kadota, T. Ohata and N. Sharkhuu
Frontier Observational Research System for Global Change, Yokohama 236-0001, Japan
*Institute of Geography, Mongolian Academy of Sciences, Ulaanbaatar 210620, Mongolia
INTRODUCTION
As a consequence of global warming, mean annual air temperatures
are predicted to increasesignificantlyinthehigh
latitudes, leading to changes in the distribution and geometries
of permafrost on a decadal scale. Degradation and thawing of
permafrost is expected to occur, especially atthesouthern
boundaries of discontinuous permafrost, resulting in significant
changes in the components of land-surface hydrology such
as surface runoff, groundwater flow, soil moistures and
evapotranspiration as well as those of energy balance.
The Frontier Observational Research System for Global
Change has promoted a research program entitled “Observational
study on the water cycle in the periphery of Eurasian
snow cover/frozen ground region in relation to climate forma
tion”. The critical components of this project are to clarify and
to define the hydrological roles of frozen ground with respect
to land-surface hydrology. This year we focused on the spatial
variations of ground thermal and moisture conditions in
Mongolia at the southern boundary of Eurasian discontinuous
permafrost. On the basis of the results obtained, this paper
discusses the local distribution and thermal condition of permafrost,
and its hydrological roles.
REGIONAL SETTING OF STUDY AREA
The Tuul river basin in the Khentei Mountains, northeast d
Ulaanbaatar was selected as an area for intense observation.
The topography is characterized by a significant relief with altitude
ranging from 1300 to 2500 m A.S.L. Mean annual air
temperature (MAAT) averaged between 1992 and 2000 was
-3.5 ‘C at the Terelj meteorological station (1520 m A.S.L).
The 20 years MAATs averaged at Zuunmod (1529 m A.S.L.)
and at the Ulaanbaatar stations (1300 rn A.S.L.) were close
to -1.0 ‘C. Recent long-term air temperature observations indicate
significant warming. Originally negative MAATs have
sometimes become positive values after the late 1990s.
Mean annual total precipitation
Ulaanbaatar station was 300
mm between 1980 and 2000, and approximately 80 % of the
precipitation occurs between June and August.
In spite of such dry climatic conditions, dense forests are
extensively developed mainly on the northern slopes, while
pastures dominate on the southern slopes and on the flat
plains.
OBSERVATIONS
Observations include measurements of ground temperatures
at shallow and greater depths, of soil moisture and of hydraulic
conductivity at two locations; in Shijir valley, west of the
Terelj station and at the Nalaikh site. In the Shijir basin, hnsect
measurements were carried out on steep and gentle
northern forest slopes (NFSs) and on southern pasture slopes
(SPSs) in the rainy season of July and during the dry season
of October and November 2002. At the Nalaikh site, we
drilled a borehole reaching a depth of 30 m in early October.
The drill site is on the wide flat pasture plain (FPP). A ground
temperature profile was measured on the 1st November.
Core samples were taken at 20 cm depth intervals and used
for estimating soil water contents.
RESULTS
Shjiir valley (Figure 1)
Steep NFSs (sites A, B, C, Q and R) havelarge angular
gravels overlain by thick organic materials composed of hu-
mus and moss, while gentle NFSs (sites E and F) dominate
silt materials. Soil textures beneath SPSs were characterized
by sand-gravels overlain by thin humus. The highest hydraulic
conductivity (I00 mld) was found at steep NFS (site A).
The values on other slopes ranged from 0.1 to IO mld. Ice
lenses were visible only beneath steep NFSs only. Higher
ground water tables were found at gentle NFSs.
Ground temperatures at 1.5 m depth were low on steep
NFSs (reaching 0 ‘C at sites A, B, C, Q and R), but higher
(ranging from 5 to 6 OC at sites I, Nand 0) on SPS. Thermal
observations at greater depths in November indicated the OG
currence of permafrost below a depth of 2.4 m on NFS (site
E). On the other hand, ground temperatures at SPSs were
about 5 ‘C at the same depths, suggesting that permafrost is
absent.
67

High moisture contents, reaching 40 and 60 % at some
depths, occurred beneath gentle NFS (sites E, F and S). At
these sites, ground water tables were found at shallower
depths. On the other hand, moisture contents were less than
20 % on steep NFSs (A, B and C) and SPSs (/and 0).
Nalaikh site (Figure 2)
At the Nalaikh site, freezing reached a depth of 0.9 m on b
1st of November. The highest temperature (3.7 'C) occurred
at 3 m depth. Below this depth, ground temperature de
creaseduntiltheuppermostportion of frozen grounds was
reached. Subzero temperatures (0 to -0.2 'C) occurred be
tween 6 and 12 m depth, indicating permafrost conditions. The
temperatures of sub-permafrost layers were slightly above 0
OC.
Significant differences were found in the water contents d
the active layer, the permafrost and sub-permafrost layers,
respectively. While water contents were less than 10 % in
the active layer, those in the upper portion of permafrost were
high, varying from 15 Yo to 27 %. The water contents in the
lower portions of permafrost were from IOto 20 % and similar
to those of sub-permafrost layers.
SPATIAL THERMAL CONDITIONS OF PERMAFROST
Generally, permafrost underlies the NFSs where the incoming
local energy fluxes on the surface are reduced due b
slope orientation, forest shading and thermal properties d
substrata. Cold permafrost with a thinner active layer occurs
beneath NFSs.
Extremely warm permafrost with thick active layers sporadically
occurs beneath FPPs. According to the increasing
rate of ground temperature change estimated by Sharkhuu
(2003), the permafrost at Nalaikh site could completely thaw
during the coming 20 years.
lshikawa, M, et 31.: Spatial conditions of frozen ground and sol ...
ables water to exchange between the bottom of the active
layer and the uppermost part of the permafrost. Such permeable
permafrost would not keep the overlaying active layer
wet. As suggested in the unevenly distributed forest, this
process may affect the land-surface hydrology even within
the small areas and needs further studies with due consideration
of the differences in permeability between contrasting
sites with cold and warm permafrost.
Figure 1. Spatial pattern of thermal and moisture conditions in the:
Shijir valley in mid-July, 2002: (a) Topography and location of each
site (b) Ground temperature profiles (c) Volumetric water contents,
depths of groundwater level and ground ice.
DISCUSSION
The influence of permafrost as an impermeable layer can be
well delineated and observed in comparisons with the distribution
pattem of soil moisture in the Shijir valley. Water from
precipitations was kept in active layers, leading to high
moisture contents and groundwater levels beneath gentle
NFSs. The differences in hydraulic conductivity and topography
between steep and gentle NFSs enable subsurface water
to flow downward laterally above the permafrost table and
to stay beneath gentle NFSs, resulting in high moisture contents
only beneath gentle NFSs (Fig. 1). Dry subsurface
conditionson SPSs are likely to be dominated by vertical
water movements, because of the absence of impermeable
layers such as permafrost.
Vertical water movements also probably occur at the
Nalaikh site underlain by extremely warm permafrost with
dry active layers. Since the amount of unfrozen water in b
zen soils increases as temperature increases to zero (e.g.
Burt and Williams 1976), large amounts of unfrozen water are
contained within the permafrost. This condition probably en-
68
Figure 2. Thermal (1st November) and moisture
profile at Nalaikh site.
REFERENCES
(early October)
Burt, T. P. and Williams, P.J. 1976. Hydraulic conductivity in frozen
soils. Earth Surface Processes, 1: 349-360.
Sharkhuu, N. in press. Recent changes in the permafrost of Mongolia.
In: Proceedings of the 8th International Conference on Permafrost.
Zurich, Switzerland.

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available ~ f l ~ ~ ~
Radio wave method for the determination of the electrical properties of
permafrost in crosshole space
V.A. Isfrafov, S.O. Perekalin, A.O. Kuchmin and *A. D. Frolov
Radionda Ltd, Moscow, Russia
*Scientific Council on Earth Ctyology, Russian Academy of Sciences, Moscow, Russia
Geocryology studies are important during the construction
and exploitation of engineering objects in permafrost areas.
Investigations of inter-borehole space by the radio
wave geo-introscopy (RWGI) method, whereby resistivity
and dielectric permittivity are simultaneously defined,
make it possible to ensure higher reliability when determining
the forms and disposition of the geological inhomogenities
within the volume under study (Borisov et al.
1993). RWGl measurements could be a reliable instrument
for the delineation of the frozen - thawed sediment
boundaries. This suggestion is made, first of all, because
of the high dependence of formation resistivity and dielectric
permittivity on the aggregate status of porous water.
The first tests in this domain were carried out in 1940
near lgarka (Eastern Siberia). These experiments have
shown (Petrovsky and Dostovalov 1947) that the propagation
of radio waves (up to frequency - 6.5 MHz) is quite
possible through about 10 m of permafrost. Since this time
there have been many tests using various radio wave
techniques to study frozen formations. It needs to be
noted that until recent years there was little evidence of a
concrete technology (special equipment, procedures for
field measurements, data processing, mode of result
presentation etc.) to make such studies.
The physical-geological basis of the RWGl method is
the relationship between the intensity of radio energy absorption
along the wave path and the electrical characteristics
of the soils. The coefficient of radio wave absorption
(k") depends on the electromagnetic field frequency and
the electrical properties of the medium:
I
o =2II f, f is the operating frequency, p and E are,
respectively, the absolute magnetic and dielectric
permittivities, p is the resistivity of a medium.
In soils with a resistivity 400 am, the radio wave absorption
is practically not dependent on the dielectric permittivity
(the medium is a quasi-conductor).
In frozen soils with high resistivity (p -hundreds to thousands
Qm) and E ~30, radio wave absorption at
frequencies of 30 MHz and higher depends considerably
on permittivity (the medium is quasi-dielectric). When
measurements are performed in an inhomogeneous medium,
k" is an integral value that summarizes all the local
changes in the medium absorbing properties along the
wave path, or more correctly, within the most essential
space for radio wave propagation (the first Fresnel zone).
Having performed measurements at two fixed frequencies,
it is possible to calculate the effective magnitudes of electric
resistivity pe and dielectric permittivity &e of soils within
the studied intervals.
A special digital borehole radio equipment of the
RWGI-2 series was designed and manufactured by "Radionda
LTD" for measurements of the intensity of the electromagnetic
field emitted using an electric (or magnetic)
dipole antennae at the fixed frequencies of 0.061, 0.156,
0.625, 1.25, 2.25, 10.0 and 31 .O MHz. Cross-borehole
measurements are usually performed following the "tomographic
survey" technique, where the field is measured at
each position of the transmitter in one borehole, over the
whole working range of receiver displacement in the other
borehole, - "in fan manner" (Fig. 1). To exclude the antenna
effects of the cable, the borehole units are connected
to the cable through dielectric inserts (coupler) with
a fiber-optic channel and an optoelectronic converter. Besides
measuring the electric (or magnetic) field intensity at
the reception point and those of the current in the transmitting
antenna, the equipment design makes it possible
to execute transmitter tuning at each position in the borehole
and to control the emission power. This makes it
possible to operate in boreholes with distances of
hundreds of meters from each other even with a short
antennae. While processing, the measured data is corrected
to exclude any differences in power emission. In
the upper part of Figure I the scheme of RWGI-CB measurements
along 6 sections with 4 boreholes (1-4) is presented
as an example. The contour of the first Fresnel
zone is shown for each section (between any pair of
holes).
To obtain reliable final data, it is essential to secure a
69
sufficient overlapping of these zones. To do this requires a
preset and suitable operating frequency for the given con-

~~~~~~ CIPEI:
UT
ditions for RWGI-CB. All this ensures the most complete
and uniform studies of inter-hole space because every inhomogenity
in the medium is explored at various angles.
~ i ~ wf i ~ - ~ ~
es ~~~~~~~
lstratov, V.A. @t al.: Radio wave method for the determination ...
G%rn.l m k %
d”
each borehole. In is evident that by means of RVGl it is
possible to extrapolate high temperature zones in an interhole
media as areas of low resistivity. The most conductive
zone corresponds to a leakage channel.
The briefness of this paper does not permit the discussion
of the very interesting results of RVGl monitoring and
the measurements of dielectric permittivity obtained at the
same site (Snegirev et al. 2003).
The modern technology of RWGl measurements and
associated data processing provides geocryological investigations
with an appropriate resolution for engineering
purposes.
REFERENCES
Borisov, B.F., Istratov, V.A. and Lysov, M.G. 1993. Method of radio
wave cross-borehole geointroscopy. Patent of Russia 2084930.
Petrovskii A.A. and Dostovalov B.N. 1947 First experience of radiowave
propagation in permafrost media. Proceedings of the
Permafrost Institute 5: 121-160. Moscow: USSR Academy of Sciences
press.
Snegirev, A.M., Velikin, S.A., Istratov, V.A., Kuchmin, A.O. and Frolov,
A.D. 2003. Geophysical monitoring in permafrost areas. Proceedings
of the 8th International Conference on Permafrost, Zurich,
Switzerland.
Figure 2. Example of the geoelectric section of effective resistivity
(fragment of the 3D map) obtained from the field studies made by the
RWGI-CB technique at the site near diamond pipe “Udachnaya”.
As an example let us consider the investigations performed
in order to delineate subsurface leakage channels
from the Sytykan water reservoir (Yakutia). The crosshole
RWGl measurements were provided using 1.25 MHz. A
fragment of the geoelectrical map is shown in Figure 2.
The section is orthogonal to the shoreline of the Sytykan
reservoir. The temperature curves are also shown near
70

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Effects of volumetric ice content and strain rates on shear strength ant 1
creep rate under triaxial conditions for frozen soil samples
M. M. Johansen and *l. U. Arenson
Department of Civil Engineering, BYGDTU, Technical University of Denmark, Lyngby, Denmark
*Institute for Geotechnical Engineering, Swiss Federal lnstitute of Technology, Zurich, Switzerland
INTRODUCTION
Rock glaciers are special geomorphological phenomena d
mountain permafrost. In some rock glaciers that have been
investigated recently, a clear shear horizon has been de
tected (Arenson et al. 2002) in which most deformation takes
place. Within an alpine environment, rock glaciers may have
temperatures close to 0 "C(Vonder Miihll et al. 2003), which
make them very sensitive to global warming. Even though
rock glaciers in the Swiss Alps have been investigated intensively
since the 1970s (e.g. Barsch 1996), there are still a
lot of unanswered questions. In particular, the loss of strength
of the frozen geomaterial, which might lead to stability prob
lems of steep slopes, is expected as an effect of a warmer
climate and is of major concern in Switzerland.
Therefore, a set of triaxial creep and shear tests were
performed to study the effect of the volumetric ice content under
natural stress conditions.
TESTS
The goal for this project was to examine mechanical re
sponse of various ice-solid mixtures from triaxial shear and
creep tests. A programme was set up to test the strength under
different confining pressures and for a range of strain rates
as well as volumetric ice contents. The material was prepared
as closely as possible to the natural conditions found in
alpine rock glaciers (e.g. Arenson et al. 2002, Vonder Muhll
et al., 2003) and tested at equivalent temperatures. To obtain
controllable and repeatable tests, the original soil was mixed
in the laboratory. Triaxial shear tests and creep tests were
carried out in a cold room with a fixed temperature at the Institute
for Geotechnical Engineering at the Swiss Federal Institute
of Technology in Zurich. Artificially-made samples with
an ice content of 30, 50, 80 and 100% by volume, respectively,
were tested under changing testing parameters. The
confining pressures s3 were 50, I00 and 200 kPa and the
strain rates used for the shear tests were 0.25, 2.5 and
25 mmlh. All tests were carried out at a temperature of about
-1 S"C. During a test, the temperature variations were about
+/-0.05 "C.
TEST RESULTS
for
Figure 1 shows the peak and the residual shear strength
samples with different ice contents at three different strain
rates. Figure 2 shows the peak and residual shear strength
for samples with different strain rates at four different ice contents.
The tests indicate that the strength of the frozen soil increases
rapidly at the beginning of the test, probably due b
icestrengthening,after which itreachespeakstrength and
drops to a lower value due to bonds cracking between Ute
soil and the ice. The authors observed no clear trend between
the strength and the confining pressure within the range d
confining pressure tested, especially for ice-rich soils. A trend
was recognized for soils with lower ice contents (30%).
Soils with low volumetric ice content and high confining pressure
showed a behaviour similar to unfrozen soils. Soil
strengthening takes place due to structural hindrance of the
solid particles after an initial slight strengthening of the ice with
increasing solid content. This results in an increase of the
strength during the test with increasing axial strain, which has
also been described by other authors (e.g. Ladanyi 1981,
Ting et al. 1983).
t
I
0- i
0.1 1 10 100
Figure 1. Shear strength (s1-s3) versus strain rate for different ice
contents. Confining pressure $3 = 100 kPa.
71

Johansen, M.M. et al.: Effects of volumetric ice content and strain rates un shear strength ...
E 3000
with four different ice contents. To observe trends better,
s
samples with different ice contents should be prepared and
" 2500
tested. In addition, the influence of the air in the sample and
0
2000
the difference between peak strength and high strain rates
should be examined.
1500
The reason for testing artificially-made samples instead d
1000
original samples extracted from a rock glacier is that it should
be possible to carry out repeatable tests. This is often not
possible when original samples are used, because the composition
of the sample can vary and this not well recognized
0 20 40 60 80 100 120
until after the test. Furthermore, the recovery of original per-
Volumetric ice content mafrost samples is very expensive due to the costs of drilling
[" +a. .25, peak . rn 2.5. peak ~ 25, peak b0.25, res. o 2.5, res. 125, res] and fieldwork. However, it was the intention to prepare and
test the samples under similar conditions found in a real rock
glacier. Further investigations on artificial soil samples may
help in developing a better understanding of rock glacier dynamics.
There is a clear trend that higher strain rates result in
higher shear strengths. In general it is found that the strength
increases with decreasing ice content except for the samples
REFERENCES
with 100% ice, which showed the highest strength (Fig. 1). Arenson, L.U., Hoelzle, M. and Springman, S.M. 2002. Borehole
The authors assume that this is because a different mecha- deformation measurements and internal structure of some rock
nism is in play. The sample showed a brittle behaviour with a glaciers in Switzerland. Permafrost and Periglacial Processes
sudden loss of strength after a peak value was reached. 13: 1 17-135.
Because of limited time, only a few creep tests were car-
Barsch, D. 1996. Rockglaciers: indicators for the present and
former geoecology in high mountain environments.
ried out. These could not be used for proper analysis. Further Springer series in physical en vir0 nmen t 16.
tests should be carried out to examine the creep behaviour d
the frozen soil in more detail.
Figure 2. Shear strength (m-03) versus volumetric ice content at
different strain rates. Confining pressure o3 100 kPa.
CONCLUSIONS
The following conclusions can be drawn from the tests on b
Zen soils presented.
Samples have to be prepared very carefully in order b
achieve repeatable tests. The ends of the samples should be
cut precisely and parallel to each other so that the stress is
equally distributed on the surface. The technique used for
sample preparation has proved to be appropriate, but an air
content of about 4% was measured. To avoid air in the Sample,differentmethods
for thesamplepreparationshould be
considered.
The strength and the mechanisms of frozen soil are very
dependent on the temperature. Lower temperature leads b
higher strength. Small temperature differences in the soil
samples and in the cold room used for tests can explain the
different values for strength found for similartests.On the
other hand, small differences in the volumetric ice content
might also have influenced the strength of the sample.
Due to the time limitations of this diploma project, it has not
been possible to examine everything in depth. In additions,
questions have arisen during the testing of the samples considering
mechanical behaviour of frozen soil which are worth
investigating further. The volumetric change in the samples
should be examined and a better method to measure the volume
of the sample should be developed.
The ice content versus shear strength should be examined
in more detail. The tests were only made on samples
72
Ladanyi, B. 1981. Mechanical behaviour of frozen soils. Mechanics
of Structured Media B: 205-245.
Ting, J., Martin, T. and Ladd, C. 1983. Mechanisms of strength for
frozen sand. J. of Geotech. Eng. 109: 1286-1302.
Vonder Muhll, D. Arenson, L.U. and Springman, S.M. 2003. Temperature
conditions in two Alpine rock glaciers. Proc. 8" Int.
Conference on Permafrost, Zurich: this issue. Lisse: Swets &
Zeitlinger.

Evolution of Alpine permafrost under the impact of climatic
fluctuations (the results of numerical modeling)
M. Kasymskaya, *D. Sergueev and N. Romanovskii
lomonosov Moscow State University, Faculty of Geology, Geocryology Department, Leninsky Gory, Russia
*Geophysical Institute, University of Alaska, Fairbanks, Alaska, USA
Alpine permafrost is characterized by considerable
spatial and temporal variability, especially near its southern
boundary. To assess the evolution of alpine permafrost
we used a 2-D computer-based model developed by
G.S. Tipenko. This model takes into account (1) long-term
climatic fluctuations, (2) the geometry of mountain relief,
(3) the thermal properties of rocks and their moisture
content and (4) the character of permafrost altitudinal
zonality.
This model was applied to study permafrost evolution
within a typical mountainous area in Southern Yakutia
(Chul'man Depression) with its extremely continental climate,
flat-topped mountains and a discontinuous distribution
of permafrost. This area has an inverse type of altitudinal
zonality, Le., mean annual temperatures of both air
and rocks rise as the absolute height increases. A profile
across a mountain massif with flat-topped divide (950 m
a.s.1.) and depression (750 m a.s.1.) was selected for
modeling. (Fig.1)
i.0
I I I I I
0 14 25 36 63 74
Figure 1. Chart of the model field
The new palaeogeographical scenario developed by
A.V. Gavrilov for the last 120 ka was used (Fig.2) This
scenario takes into account the effect of permafrost altitudinal
zonality on the dynamics of mean annual ground
temperatures. It was assumed that ground temperatures
on the tops of plateaus (divides) rise above 0°C in the periods
of maximum climatic warming.
Temperature, "C
-12 -10 -8 -6 -4 -2 0 2 4
1- on the mountain top -2- on the valley bottom
Figure.2. The curves of mean annual ground temperatures in accordance
with the pako- geographic scenario.
1
Numerical modeling showed that in the given type of
mountains, permafrost temperatures in depressions and
river valleys are substantially higher than those on watersheds.
In periods of climatic warming, taliks can develop
on the tops of plateaus. Thus, infiltration of atmospheric
precipitation becomes possible in summer periods. It was
also found that the maximum depth of permafrost is
reached, with some time lag, after the minimum permafrost
temperature has been attained.
The results of this modeling make it possible to estimate
the effect of climatic fluctuations on the extent and
depth of permafrost, as well as the amount of increase of
groundwater through talik formation on local divides during
phases of climatic warming. The appearance of these
taliks considerably effect the groundwater and hydrological
regime of local rivers, especially in the winter period.
73

Kasymskaya, M. et ai.: Evolution of Alpine permafrost under the impact..
- 1 I
300 ' I
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
1- surface of relief Number of blocks
-2- permafrost lower boundary, 83 Kyr B.P.
-3- permafrost lower boundary, 6 Kyr B.P.
-4- permafrost lower boundary, recent time
1
""
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
_-"x 1- surface of relief Number of blocks
-5- permafrost lower boundary, 105 Kyr B.P.
-6- permafrost lower boundary, 22 Kyr E.P.
Fig.3. Surface of relief (upper boundary of alpine permafrost model)
and position of permafrost lower boundary in different time periods.
REFERENCES
Sergueev,D., Tipenko, G., Romanovskii, N., Romanovsky, V., and
Kasymskaya, M. 2002. Impact of Mountain Topography and Altitudinal
Zonation on Alpine Permafrost Evolution and Ground Water
Hydrology, AGU 2002 Fall Meeting Abstracts, San Francisco,
California, USA.
Sergueev,D., Tipenko, G., Romanovsky, V., and Romanovskii, N.
Mountain permafrost thickness evolution under influence of longterm
climate fluctuation (results of numerical simulation). Proceedings
of the 8th International Conference on Permafrost (In
press).
74

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
An estimate of organic carbon inputo the Arctic seas from permafrost and
soils of the East Siberian coasts
A.L. Kholodov, D. E. Fyodorov-Davidov, E. M. Rivkina, S. V. Gubin, V.A.Sorokovikov, L.A. Fominikh, and D.A. Gilichinsky
institute of Physicochemical and Biological Problems in Soil Science, Russian Academy of Science, Pushchino, Russia
Permafrost is a significant reservoir and potential source
of ancient organic matter (plant remains, humified organics,
etc.) and greenhouse gases (C02 and CH4). Due to
the degradation of permafrost as a result of coastal erosion,
thermokarst formation and thawing of submarine
permafrost, this organic carbon is easily released and reabsorbed
again into the present biogeochemical cycle.
The task of this study is to estimate possible organic input
into the Laptev and East Siberia seas and atmosphere
due to these processes as one objective of the IPA Arctic
Coastal Dynamics program.
Investigations were carried out at three key sites: Bykovsky
peninsula (129030'E, 71040'N.), Cape Svyatoi Nos
(14O00'E 72055'N), both in the Laptev Sea region, and
Cape Chukochyi (159-55'E 70005'N), East Siberian Sea
coast. All sites are characterized by the widespread occurrence
of fine-grained Quaternary sediments with large
quantities of organic matter and biogases. There are two
main types of landscapes associated with the deposits.
One is the late Pleistocene (I-alQIII) accumulative plain
(Edoma formation or ice-complex) composed of syncryogenic
ice-rich sediments. Mean annual ground temperatures
within this landscape are -11 to -14OC. Shore cliffs
are 20 to 40 m in height. The other type of landscape is
characterized by alas depressions resulting from the
thawing of ice-complex under thermokarst lakes. The alas
complex includes layers of Holocene lake-boggy deposits
(I-bQlv); layers of tabular deposits (thawed and redeposited
ice complex; tbQlll-lv), and residual ice complex
(Fig. I). Mean annual ground temperature is -9OC and the
height of Laptev shore cliffs is less than 10 m but can
reach 20 m along the East Siberian coast.
One of the main processes in this region is coastal
erosion. The length of eroding shores of the Laptev Sea is
1000-1300 km, and for the East-Siberian Sea up to 2000
km. The rate of coastal retreat is 2-6 mlyear (Are 1998).
According to data obtained for the Cape Chukochyi at the
East Siberian Sea, the coast has retreated at an average
rate of 4.5 mlyear during the past 18 years. Another im-
portant process which takes place in the region is erosion
of the sea floor. This process is most active on the shallow
shelves near the coast and on the seabed of former is-
lands that were destroyed by erosion over the past centuries.
Total content of organic carbon was determined on dry
soil samples using the wet oxidation method. Volumetric
total Corg content for massive frozen soil was calculated on
the basis of an assumed soil density (1700 kglm3). Calculations
were carried out using a volumetric ice content
of 75% for the ice complex and of 50% for the alas complex.
w C org, YO W C org, %
A 0 2 4 048lZ El 0 2 4 O B *
Figure 1. Texture, water and carbon contents of ice complex deposits
(A) and alas complex (B) on Bykovsky Peninsula.
Supply of Corg for the different relief forms was determined
for the active layer at the CALM sites R13, and
R13A (Brown et al. 2000). Thickness of the active layer
was taken as 0.5 m for hummocks, 0.4 for spot medallions,
0.3 m on the slope of hummocks, and 0.1 m in
cracks. Soil density was assumed as 150-200 kglm3 for
organic horizons, 800 kglm3 for mineral horizons in the
upper part of the profile, and I700 kglm3 for the lower part
(Fig. 2). Results of the calculation of total Corg in frozen
deposits and modern soils are given in Table 1.
It was determined that 1 m3 of ice complex contains 5-
12 kg of Corg Due to erosion of 1 m2 of coastal territory, an
amount of 25 to 425 kg of Corg can be delivered into Arctic
seas.
Deposits of the alas complex are characterized by a
similar total Corg (8-12 kglm3). The main part of organic
75
a
16 '$

Khuladav, A. L. et 31.: Estimate of organ carbon input, Arctic seas. East Siberian coasts,,,
supply (75%) in these deposits is concentrated in upper On the widespread marine terraces, river deltas and
parts of the section which are composed of lake-bogy alas depressions of Northeast Russia, where the bluff
sediments. Tabular deposits have less organic matter
relative to the initial ice complex. Average organic content
in a 50m thick section of massive ice complex on the Bykovsky
peninsula is 22 kglm3, but a 10 m section of tabular
deposits contains 3 to IOkglm3. Thus, due to thawing of
the ice complex during development of the thermokarst
lake, decay of up to 50% of the organic matter took place
(Fig. 1).
height is not more than 5 m, the total Corg input to the Arctic
seas from modern soils is comparable with the Corg
from cliffs. Total amounts of Corg, delivered to the Arctic
seas due to coastal erosion of the Laptev and East Siberian
seas can be up to 0.8 million tonslyear or more, including
0.01 million tons of modern vegetation, 0.03 and
more of modern soil, and more than 0.75 buried in permafrost.
These values are comparable with the input from
river discharge (0.85 million tonslyear) (Danyushevsky
1990).
Due to the erosion of marine banks, the organic carbon
could also be released into the atmosphere. The concentration
of CH4 in permafrost varies from 2 to 40 mllkg and
of C02 from I to 20 mllkg; it should be noted that catbon
dioxide is ubiquitous while the occurrence of methane is
determined by the genesis of the deposits and by the type
of cryogenic stratum. For example, the ice complex on the
East Siberian coast is free of methane, while the alas
complex contains 20 to 30 mllkg of CH4. We have calculated
the amount of methane which could be released after
thawing and abrasion of the Cape Chukochyi alas outcrops
(with sediments containing 30 ml CH4lkg and
outcrop sizes being up to I km in length, - up to 20 m in
Figure 2. Typical profile of modern soil (Cape Chukochyi).
height and with coastal retreat amounting to -1 m). The
resulting value is close to 0.4 tonnes.
Table 1. Content of organic carbon in frozen deposits and modern soil
of Laptev and East Siberian coastal zone.
Borehole Deposits Average
ACKNOWLEDGEMENTS
TOC, kglm3
Bykovsky peninsula
These investigations were carried out as a part of Russian
1-01 I-alQIII 11.2
Academy of Sciences “World Ocean” program and sup-
2-01 N-l-bQIII-lv 12.2
ported by the Russian Foundation for Basic Research
Svyatoy Nos cape
(2003). The modern soil depths of thawing and organic
3-01 I-aIQ~~.~~~ 5.5
matter content were studied within the framework of the
4-01 QII-III 3.4
Chukochyi cape
NSF-supported CALM project (Brown et al. 2000).
2-91 mQI 71 .I
3-91 IQIII-IV 8.2
4-91 mQtl-rll 8.6
5-91 IQIII-IV 11.2
REFERENCES
24.7
Are F. 1998. The contribution of shore thermo-abrasion to the Laptev
Sea sediment balance. In Proceedings of the 7th Intern. Confer-
10-91 I-alQIII 5.9
ence on Permafrost. Yellowknife, Canada: 25-31,
Modern soil (Chukochyi cape)
Brown J., Hinkel K.M. and Nelson, F.E. 2000. The Circumpolar Active
Edorna Hummock 19.6
Layer Monitoring (CALM) program: Research designs and initial
Spot medallion 14.8
results. Polar Geography 24 (3): 165-258.
Hummock slope 24.5
Danyushevsky A. (ed) 1990. Organic substance of bottom sediments
Crack 5.6
in polar zone of World Ocean. Leningrad, “NEDRA (in Russian).
Alas Hummock 21.7
6-91
8-91 1Q11t.n
IQIII-IV
11.0
Spot medallion 25.3
Hummock slope 23.1
For the modern soils of the Edoma surface of Cape
Chukochyi, the content of Corg is in the range of 5-9 kglm3.
For alas depressions it is 7.5-11.5 kglm3. Similar values of
Corg for sites along the Bering Sea coast (Svyatoi Lavrenty
Cape) were obtained by D. Zamolodchikov (see abstract
in this volume for site discussion). The average supply of
total Corg in modern vegetation is about I kglm2.
76

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available l n ~ o ~ ~ ~ ~
Long-term monitoring of borehole temperatures and permafrost-related
data for climate change research and natural hazard management: examples
from the Mattertal, Swiss Alps
L. King, R. Hot T. Hen and *S. Gruber
lnstitut fur Geographie, Justus Liebig-Universitat Giessen, Germany
"Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
INTRODUCTION
Mattertal and neighbouring Saastal are certainly amongst the
best investigated areas in the Alps with regards to permafrost
distribution and natural hazard management. An extremely
steep topography and traffic routes to touristic sites (e.g.,
Zermatt, Saas Fee) have led to intense research efforts by
transport authorities and scientists in order to minimize risks
for the population and tourists and to acquire the data necessary
for natural hazard assessment and protection. Moreover,
long-term temperature monitoring in deep boreholes and
continuous climate measurements provide valuable data for
permafrost distribution modelling and climate change research.
BOREHOLE MONITORING
RESEARCH
FOR CLIMATE CHANGE
Three boreholes have provided us with significant data concerning
subsoil rock temperatures. The locations and the
characteristics of the two boreholes on the Stockhorn plateau
and the borehole at Ritigraben are shown in Table 1. The
temperature curves vary considerably, mainly due to their
respective positions and surface structures. Permafrost thickness
is expected to be approximately 170 meters (Stockhorn
Plateau) and more than 30 m (Ritigraben), respectively. The
temperature curves show evidence of climatic warming during
the last decades, however, the separation of topographic
from transient effects in complex topography is a challenging
task. The two boreholes at the east-west running crest of the
Stockhorn show a very strong effect of exposure to permafrost
characteristics. On the plateau and close to the northem
face of the ridge -2,5" C have been measured at 30 metres
depth, however, only -0,9" C were recorded at the southern
face.
The Stockhorn and Ritigraben boreholes provide data for
long-term monitoring of permafrost temperatures, an important
source of data for climatic change research. In the high
mountain areas of Europe, this idea was initiated by the E U
project PACE (Permafrost And Climate in Europe) and the
network is maintained by its ESF-sponsored continuation,
PACE-21. The data is also integrated into international or na
tional databases as GTN-P or PERMOS (PERmafrost
Monitoring Switzerland), forming a basis for global comparative
research.
CLIMATE MONITORING AND PERMAFROST
MODELLING
Long-term data gathered by the Inter-cantonal Measurement
and Information System (IMIS, ENET) for avalanche warning
provides permafrost-related information such as air and
ground temperatures, precipitation, snow depth, global radiation,
wind speed and wind directionintheMattertal.These
stations are ofien located in permafrost areas, as the 0°Cisotherme
in the Mattertal lies slightly above 2600 m a.s.1.
(Tab. 2).
In connection with information taken from DTMs, aerial
photography, satellite imagery and research undertaken at the
Giessen Institute for Geography (Gruber 2000; Philippi et al.,
in press), this climate information forms an excellent database
forfurtherpermafrostmodelling (Gruberand Hoelzle 2001)
using a Geographical Information System (GIs). A field
check with BTS measurements was done in March 2003. It
is intended to implement this data into a neuronal network as
this could prove to be a valuable tool for a better modelling d
permafrost.
MONITORING FOR NATURAL HAZARD
MANAGEMENT
The strong variability of climatic parameters in this relatively
dry andcontinentalclimate may lead to extreme climatic
events that could trigger catastrophic mass wasting
processes due to the extremely steep topography of the
whole Valais region. The increasing frequency of natural hazard
events are a matter of great concern to the local, regional
and national authorities. Great attention is therefore paid to the
assessment and management of natural hazards.
More and more, it is realized that permafrost plays an important
direct or indirect role in the assessment and prevention
77

of these hazards (Hen et al. in press). The responsible
authorities and research organizations therefore support ap
plied research of the mudflow origins in permafrost areas and
the origin of unexpected water releases, the installation d
early warning devices in mudflow canyons, the monitoring d
degrading permafrost and the examination of the stability d
avalanche protection structures in permafrost areas.
Table 1. Borehole Stockhorn Plateau and Ritigraben, location and
borehole characteristics.
Stockhorn Ritigraben
Co-ordinates
Altitude
Slope inclination
Date of Drilling
Borehole depth
Permafrost thickness
Active layer depth
ZAAIMAGT
Active layer
45"59'00" N
07"49'05" E
3410 m a.s.1.
8" S
August 2000
100 m I30 m
170 m (approx.)
3.1 m
17.7 m/-2,5"C
bedrock
Temperature [ i C]
King, L. et al: Barehole temperatures and permafrast-related data.,.
46" 10' 27" N
07" 50' 56" E
2615 m a d.
14" NW
October 2001
30 m
>30 m
3.8 m
14 m/-0,37"C
boulders
-10 -8 -6 -4 -2 0 2 4 6 8 10
0
Table 2. Selected IMlS and ENET-Climatic stations used for modelling
purposes in the Matter valley.
Name
2001
Altitude MAAT
Gornergrat snow station 2950m -2,l "C
Gornergrat wind station 31 30m -2,7"C
Saas Platthorn 3246m -3,9"C
Saas Schwarzmies 281 Om -1,3"C
Saas Seetal 2480m +0,5"C
St. Niklaus Ob. Stelli Glacier 291 Om -1,6"C
Zermatt Platthorn 3345m -4,5"C
Zermatt Triftchumme 2750m -0.6"C
CONCLUSIONS
Long-term monitoring of permafrost temperatures in deep
boreholes and the analysis and interpretation of climate data
forms a valuable source of information for climate change re
search and permafrost distribution modelling. This information
can also be used for natural hazard detection and protection.
10
REFERENCES
- e
20
30
40
1 50
60
70
80
, , -
. . . . . . . . .
, ,
Maximum temperature
Gruber, S. 2000. Slope instability and permafrost - a spatial analysis
in the Matter valley, Switzerland. - unpublished diploma
thesis, Institute for Geography, University of Giessen, Germany.
Gruber, S. and Hoelzle, M. 2001. A statistical modelling of mountain
permafrost distribution: a local calibration and incorporation of
remotely sensed data. Permafrost and Periglacial Processes 12
(1): 69-77.
Hem, T., King, L. and Gubler, H. (in press). Microclimate within
coarse debris of talus slopes in the alpine periglacial belt and
its effects on permafrost. 8th International Conference on Permafrost,
Proceedings, Zurich, 19-25 July 2003.
Philippi, S., Herz. T. and King, L. (in press): Near-surface ground
temperature measurements and permafrost distribution at Gornergrat,
Matter valley, Swiss Alps. - Poster Abstract. - In: International
Conference on Permafrost, Proceedings, Zurich, 19-
25 July 2003.
90
100
-0,005 0 0,005 0,Ol 0,015 0,02 0,025
Temperature gradient [ iCh]
Figure 1. Rock temperatures and thermal gradient of the 100 meter
deep Stockhorn borehole (01-10-2001 to 30-09-2002).
An early warning device has been installed in the Ritigraben
gully above St. Niklaus. In the event of a debris flow it
would automatically cause the main road in the valley to be
closed. In a further step it is intended to use precipitation data
from the weather station
the Ritigraben borehole as an early
warning parameter, that is located close to the starting point d
the debris flows.
78

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available 1~~~~~~~~~~
Destruction of coasts on the Yugorsky Peninsuland on Kolguev Island,
Russia
A. 1. Kizyakov and * D. D. Perednya
Earfh Cryosphere lnsfitufe
SB RAS, Tyumen, Russia
*Moscow State University, Faculty of Geography, Russia
Cryolithological, environmental and climatic features which
determine coastal processes were investigated at the western
coast of the Island Kolguev (Barents Sea) in summer 2002
and at the central part of the Yugorsky Peninsula coast (Kara
Sea) in the summers of 1999,2001 and 2002 (Cherkashov et
al. 1999, Kizyakov et al. 2002, Perednya et al. 2002).
Coasts at both sites consist of frozen sandy-clayey deposits
enclosing tabular (massive) ground ice. Such tabular ice
bodies control the formation of thermodenudational cirques as
characteristic landforms: such thermocirques develop in
slopes where tabular ground ice is found near the surface.
They are formed by a variety of processes: thermokarst,
coastal and lateral thermoerosion, retrogressive thaw slumps
and earth Rows.
The observed key sites d&r in the activity of hydrodynamic
processes. The western coast of Kolguev Island is
open to wave action and the dynamically active ice-free ps
riod at the Barents Sea is longer than in the Kara Sea. At the
Yugorsky key site, the hydrodynamic processes are weaker
as a result of a very short ice-free period: the Kara Sea is
completely covered with ice from October to May. Drift ice
stays practically all summer. These conditions determine the
character and dynamics of coastal destruction at the study
sites.
Coastal bluffs on Kolguev Island are up to 40-50 m high.
The coasts are of thermoerosion type, directly affected by
waves. Bluffs are steep to overhanging. At the slope foot
wave-erosion niches are formed and often remain partially
filled with snow. The beach is narrow, up to 10 m wide and
the bluff edges are dissected by numerous gullies which reflect
a polygonal pattern (Fig. I),
The three investigated coastal thermocirques are 200-400
m long (along the coast), and up to 200 m wide (landward).
Analysis of the shoreline is done based on field inspection
and aerial images from 1948 and 1968. Coastal retreat averages
3-4 mlyear (Perednya et al. 2002).
One of the thermocirques formed after 1968. The retreat
rate of its scarp was therefore no less than 5 mlyear. Older
thermocirques, however, are much less active and most
likely retreat at a rate of less than1 mlyear. At the Yugorsky
key site, coastal bluffs have relative heights from 12 up to 29
m. Coasts are of the thermoerosion-thermodenudation type
narrow beach (I), s~ee~ to ovefh
w~ve-erod~d ni~h~s (3) at the slope f~~~ are
urnerou~ ~ullies at the b i ed ~ ~
~a~t~rns.
and wave-erosion niches are not formed (Fig. 2). A complex
of thermodenudation slope processes destroys coastal bluffs.
Sediments are washed away by tides and most actively by
storm surges after having been deposited on a beach. These
sediments are involved in a shore-parallel drift directed westtoeast.
In 2001, a monitoring network was established at a
thermocirque scarp edge. The dynamics of long-term coastal
retreat and thermocirque growth was recorded based on a
comparison of field and remote-sensing data. Interpretation d
aerial images from 1947 was combined with an analysis d
topographic maps compiled using images from 1967.
The topographic survey of thermocirques and coastal-bluff
edges was performed by theVNIIOoceangeology Institute,
St-Petersburg in 2001 (Kizyakov 2002, Kizyakov et al.
2002). The scarp most actively retreating from 1947 to 2001
was analysed and the retreat rate determined as 0.6 to 1
mlyear, while the coastal-bluff retreat amounted to 1.3 - 1.6
mlyear. The topographic survey enabled the compilation of a
digital terrain model in SURFER software, which provided a
reconstructed topography of the area prior to thermocirque
formation and formed the basis forcal culating the sediment inputfromthe
thenocirque in 54 years. According to these
calculations (Tab. I), the thermocirques on the Yugorsky
79

coast increase the sediment yield, because the scarp edges
are curved and thus longer than the shoreline for the smoothfaced
bluffs.
Kizyakov, A. I. et al.: Destruction of coasts an the Yugorsky Peninsula
ACKNOWLEDEMENT
The study was performed within the framework of INTAS the
projects, grants 01-2329 and 01-2211.
REFERENCES
The corresponding sediment flux from a vast catchment area
is concentrated through a narrow exit. On average, sediment
input from thermocirques appeared to be approximately 3
times as high as from the smooth-faced coastal bluffs. The
total input of thermocirques for a large portion of the coast is
rather low because the latter occupy only about 6% of the
Yugorsky shoreline.
Cherkashov, G.A., Goncharov, G.N., Kizyakov, A.I., Krinitsky, P.I.,
Leibman, M.O., Persov, A.V., Petrova, V.I., Solovyev, VA., Vanshtein,
B.G. and Vasiliev, AA. 1999. Arctic coastal dynamics in
the areas with massive ground ice occurrence. Report of an
International Arctic Coastal Dynamics Workshop, Woods Hole,
2-4 November 1999. Geological Survey of Canada Open File
3929.
Kizyakov, A.I., Perdnya, D.D., Firsov, Yu.G., Leibman, M.O. and
Cherkashov G.A. 2002. Character of the costal destruction and
dynamics of the Yugorsky Peninsula coast. Arctic Coastal Dynamics
3rd Workshop Proceedings, Oslo, 2-5 December 2002,
(in press).
Kizyakov A.I. 2002. The forms of Kara Sea coasts destruction in
conditions of widespread tabular ground ice. Extreme phenomena
in cryosphere: basic and applied aspects. Abstracts of
the International Conference, Pushchino, 12-1 5 May 2002.
Pushchino: ONTl PNC RAS.
Perdnya, D.D., Leibman, M.O., Kizyakov, A.I., Vanshtein, B.G. and
Cherkashov, G.A. 2002. Coastal dynamics at the western part of
Kolguev Island, Barents Sea. Arctic Coastal Dynamics 3rd
Workshop Proceedings, Oslo, 2-5 December 2002, (in press).
Table 1. The rate of retreat due to coastal thermoerosion and
modenudation at the Yugorsky Peninsula and sediment yield
km of the shoreline
therper
Destructive Retreat rate Sediment yield
processes m per year m3 per year
Coastal thermoerosion 1.3-1.6 30100
and falls at the coastal
bluffs
Complex of 0.6-1 -04 89000
thermodenudation
processes in thermocirques
Thus, the western coast of Kolguev Island is of the thermoerosion
type, and collapses as a result of direct mechanical
and thermal wave influence. The coast of Yugorsky
Peninsula in its central part, however, is of the thermoem
sion-thermodenudation type and collapses mainly due b
slope thermodenudation processes. At both study sites, thermocirques
develop within smooth-faced slopes by thawing d
tabular ground ice. The coastal retreat on Yugorsky Peninsula
for the smooth-faced bluffs is slower by a factor of two than
onthewesterncoastofKolguevIsland.Activetherrnocirques
of Kolguev Island are developing up to 5 times faster
than on Yugorsky Peninsula. The development of thermocirques
in the coastal zone increases the sediment yield per
metre of a shoreline.
This points toward more active destruction of the Kolguev
site in the western aspect of the study coast of Kolguev Island
and specific hydrodynamic features (longer ice-free pe
riod, more wave action) of the Barents Sea.
80

8th International Conference on Permafrost Extended Abstracts an Current Research and Newly Available ~~f~~~~~~~~
New insights of mountain permafrost occurrence and characteristics in
recently deglaciated glacier forefields through the application of electrical
resistivity tomography
C. Kneisel
University of Wuerzburg, Department of Physical Geography, Wuerzburg, Germany
INTRODUCTION
Different geophysical methods have been applied successfully
in mountain regions to detect mountain permafrost. Since
geoelectrical methods are most suitable for investigating the
subsurface with distinct contrasts in conductivity and resistivity,
DC resistivity soundings constitute one of the traditional
geophysical methods which are standardly applied in pemafrost
research to confirm mountain permafrost over many
years. During recent years advances have been achieved in
usingthetraditionalmethodsbut with more powerful instruments
and modem data processing algorithms which enable
two-dimensional sutveys and data processing (Vonder Muhll
et al. 2001).
The study sites are situated in the Upper Engadine, eastem
Swiss Alps, in the vicinity of St. Moritz (46"30'N,
9'50'E). Numerous rock glaciers are obvious geomorphological
indicators of the discontinous distribution of alpine
permafrost. Various aspects of mountain permafrost have
been investigated in the past decades with the main focus on
rock glacier permafrost. Discontinuous permafrost is encountered
above 2400 m a.s.1. The investigation of permafrost in
glacier forefields at high altitudes in the Upper Engadine was
performed during recent years using onedimensional gm
electrical soundings as the standard geophysical technique b
detect mountain permafrost. The results of numerous measurements
confirmed the local presence of perennially frozen
ground and ground ice in the investigated glacier forefields
(Kneisel 1998, 1999).
On heterogeneous ground conditions the interpretation d
one-dimensional soundings can be difficult, as lateral variations
along the survey line can influence the results significantly.
The sounding curve produces an average resistivity
model of the survey area, but individual anomalies such as
small permafrost lenses will not show explicitly the in results.
Hence, the interpretation is limited with respect to the characteristics
and the local distribution pattern. Recently, twodimensional
electrical resistivity tomography which overcomes
this problem using multielectrode systems and twodimensional
data inversion was applied in the same glacier
forefields and at the same location confirming the assumptions
made by the one-dimensional vertical soundings.
The motivation of this contribution is to illustrate the opportunities
and potentials of two-dimensional DC resistivity b
mography for periglacial geomorphology providing a more
detailed characterization of the subsurface as compared to the
results obtained by traditional one-dimensional vertical
soundings.
METHODS
For the one-dimensional geo-electrical soundings a GGA30
Bodenseewerk instrument was used. Three-layer models
were calculated which are assumed to represent the most
probable case with respect to the local geomorphological
situation. For the interpretation of one-dimensional data the assumption
is made that the subsurface consists of horizontal
layers and that the resistivity changes only with depth but not
horizontally. The typical one-dimensional geeelectrical
sounding curve obtained on alpine permafrost shows a three
layer model with an increase of resistivity at shallow depth
representing the active layer and the permafrost underneath,
followed by a sharp decrease of resistivities greater current
electrode distances representing the unfrozen ground beneath.
For the two-dimensional electrical tomographies a SYSCAL
Junior Switch system was applied. The surveys were conducted
using the Wenner configuration.
Resistivity values of frozen ground can vary over a wide
range depending on the ice content, the temperature and the
content of impurities. The dependance of resistivity on tem-
perature is closely related to the amount of unfrozen water.
Perennially frozen silt, sand, gravel or frozen debris with
varying ice content show a wide range of resistivity values
between 5 kWm to several hundred kWm (Haeberli and
Vonder Miihlli996).
81
RESULTS
The resistivity sounding profile shown in Figure 1 was measured
in a high altitude glacier forefield using the symmetrical
Schlumberger configuration. The graph shows only a slight

Kneisel. C,: Mountain permafrost, glacier forefields, electrical resistivity tomography,.,
increase of the resistivity values. A three-layer model inversion
was performed suggesting a first layer of 3 to 7 m representing
the active layer and a second layer with a thickness
of about IOto 25 m. The resistivity values obtained for this
middle layer (between 20 km and 35 km for equivalent models),
lie in the range of perennial frozen ground and suggest a
shallow permafrost occurrence at this site. The third layer
with better conductivityrepresents the unfrozen ground underneath.
Figure 2. Resistivity tomogram of a Wenner
permafrost
survey with mountain
B
E
Figure 1. One-dimensional DC resistivity sounding profile
I
The described examples from two surveys in the periglacia1
environment illustrate that electrical resistivity tomograph
is a powerful tool to determine the structure of the subsurface
at different resolutions, depending on the electrode spacing
and array geometry employed. For a reliable interpretation d
the survey results, knowledge of the geomorphological and
geological setting is essential. Further details of applicability d
different array geometries, advantages and disadvantages d
the electrical resistivity tomography and its application b
various geomorphological studies are given in Kneisel
(2003).
REFERENCES
This shape of geoelectrical sounding graphs is considered
to be typical of permafrost in glacier forefields. Similar curves
have also been measured in other investigated forefields.
With the special conditions in glacier forefields, permafrost OG
currences can be assumed to be shallow, "warm" and thus,
rich in unfrozen water, what can explain the comparatively
low apparent resistivity values (Kneisel 1999). Figure 2 illustrates
the results of the two-dimensional resistivity survey
performed at the same location where the one-dimensional
geoelectrical soundings were performed in the previous study
which already indicated a permafrost occurrence. Thedark
grey colors represent perennially frozen ground. Through the
onedimensional sounding similar thickness of the active
layer and depth of the permafrost were obtained. However,
since one-dimensional geoelectrical soundings produce only
an average resistivity model of the survey area, the lenticular
high-resistive areas as shown in the 2-D image could not be
deduced.
Additionally, topography was incorporated in the inversion,
resulting in a more realistic interpretation of the obtained
resistivity image of the subsurface. This is especially evident
in the depression between station 45 and 55 where the active
layer is markedly thinner due to the late-lying snow. Here,
the potential of electrical resistivity tomography for geomorphological
investigations compared to onedimensional
soundings is shown clearly, as a more sophisticated interpretation
of the subsurface can be derived. Furthermore, the
opportunity for a real mapping over larger areas is a major
advantage for geomorphological studies.
Haeberli, W. and Vonder Muhll, D. 1996. On the characteristics and
possible origins of ice in rock glacier permafrost. 2. f. Geomorph.
N.F. Suppl. 104: 43-57.
Kneisel, C. 1998. Occurrence of surface ice and ground icelpermafrost
in recently deglaciated glacier forefields, St. Moritz area,
Eastern Swiss Alps. 7th International Conference on Permafrost,
Yellowknife, Canada. Proceedings: 575-581.
Kneisel, C. 1999. Permafrost in Gletschervorfeldern - Eine vergleichende
Untersuchung in den Ostschweizer Alpen und Nordschweden.
Trierer Geographische Studien 22.
Kneisel, C. in press. Electrical resistivity tomography as a tool for
geomorphological investigations - some case studies. In: L.
Schrott, A. Hoerdt and R. Dikau (eds.). Geophysical methods in
geomorphology. Z. f. Geomorph., Suppl.
Vonder Muhll, D., Hauck, C., Gubler, H., McDonald, R. and Russill,
N. 2001. New geophysical methods of investigating the nature
and distribution of mountain permafrost with special reference
to radiometry techniques. Permafrost and Periglacial Processes
12: 27-38.
82

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Evidence of paleotemperature signals in mountain permafrost areas
T. Kohl and *S. Eruber
Geophysical Institute, ETH Zurich, Switzerland
*Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
INTRODUCTION
In Alpine environments ground temperature changes may
have a significant effect on permafrost. As a consequence,
future warming is likely to lead to increased instability
in mountain permafrost terrain. However, field studies
of thaw-related slope instability are hampered by the
natural variability of geomorphic processes, both in space
and time. To quantify the impact of recent climate variation,
borehole temperature measurements were taken using
30 thermistors at the 3400 m high Stockhorn in the
central part of the Swiss Alps. The measurements show
non-linear thermal profiles, with pronounced near-surface
temperature deviations from the deeper thermal gradient.
Although climatic-induced temperature effects seem to be
evident, other topography-induced thermal signals cannot
be excluded. The present analysis aims at the investigation
of coupled transient and topography effects which
should also elucidate the significance of temperature
measurements in alpine terrains.
temperature distribution and accurate topography representation.
The analysis of these near-surface energy
fluxes is therefore represented by a two-dimensional homogeneous
model assuming constant basal heat flow at
3.5 km depth below the Stockhorn but with transient surface
temperature variation. The model included 20,000
nodes. The mesh was strongly coarsened from surface
( 4 m distance between nodes) to greater depth (100 m
distance). Topography was taken at IOm distance from a
DTM along a N-S running profile of the Stockhorn.
RESULTS
METHOD
The investigation was conducted using the FE code
FRACTure (Kohl and Hopkirk 1995). One of the features
of this code is the integration of complex topography and
surface condition into a robust and reliable numerical
code. The FE-forward scheme is now linked with the
UCODE computer program which can be broadly used for
various applications. In combination, UCODE and FRAC-
Ture perform inverse modelling, posed as a parameterestimation
problem, by calculating parameter values that
minimize a weighted least-squares objective function using
non-linear regression, With this tool complex multidimensional
inversion studies can be theoretically accomplished.
However, numerical instability and nonuniqueness
represent well-known problems in inverse modelling,
and obtaining useful results can become more difficult for
highly heterogeneous conditions. Therefore, the problem
was focused on properties that seem to have the major
impact on the near-surface temperature field: surface
Figure 1. Temperature profiles along selected sites: the measured
temperature data are given by black dots, the T-profile at x=2075m
represents the calculated temperature from steady-state assumptions.
83

only a small displacement of the borehole location would
result in completely different T(z) profiles. Under steadystate
conditions, the results are nearly independent of
thermal conductivity along the borehole depth. To account
for transient effects, a climatic time series from a nearby
study (Bohm et al. 2001) was initially superimposed on the
surface temperature. When performing 2nd order regression
on these strongly varying data, basically a I .3"C
warming between 1900 and 2000 can be identified, between
1800 and 1900 surface temperature did not change
drastically. However, our calculations have shown better
fits with a more pronounced warming in the last 50 years.
Figure 2 illustrates the additional curvature in the T(z) profile
due to climatic warming.
locations to optimize different configurations and temperature
logging devices at specific sites.
REFERENCES
Bohm R., Auer I., Brunetti M., Maugeri M., Nanni T. and Schoner W.
2001. Regional temperature variability in the European Alps 1760
- 1998 from homogenised instrumental time series. International
Journal of Climate 21: 1779-1801.
Kohl T. and Hopkirk R.J. 1995, "FRACTure" a simulation code for
forced fluid flow and transport in fractured porous rock.
Geothermics 24(3): 345-359.
CONCLUSION
Our analysis has demonstrated that the ground surface
temperature distribution is highly sensitive to the shallow
(-100 m deep) subsurface temperature field. In alpine terrains,
energy fluxes can even be totally dominated by topography
without any influence of basal heat flow. Under
these circumstances, temperature measurements need to
be carefully designed and located at those points which
are less sensitive and which provide the most useful data.
Our procedure can also be applied successfully to further
84

Hydrological modelling of (sub)arctic river systems: responses to climate
change
E. Koster, R. Dankers and S. van der linden
Centre for Geo-ecological Research, Department of Physical Geography, Utrecht University, Utrecht, lhe Netherlands
PERMAFROST HYDROLOGY
Hydrologic modelling of river regimes is restricted by our
poor knowledge of permafrost-related processes and by a
paucity of data (Woo 2000). Frozen ground is relatively
impermeable, especially when ground ice content is high.
Warm spring temperatures melt the bulk of the snowpack
within a few days or weeks, while the soils are still frozen.
The closer to the summer solstice, the greater the incoming
solar radiation so that the spring thaw is often intense
(Woo 2000), leading to overland flow as the rate of melt
exceeds infiltration capacity. When the active layer thickness
increases towards the end of the thawing season, infiltration
rates increase and surface runoff is correspondingly
reduced. Not surprisingly, the annual catchment
water balance in permafrost areas often shows a high
runofflprecipitation ratio. Typical values range from 0.7-0.8
in high-Arctic polar deserts to 0.4-0.6 in sub-Arctic tundra
regions and wetlands because of higher evaporation (Woo
2000). However, wide variations in ratios do occur between
different catchments. During summer with greater
thaw depth, depleted groundwater stores, and increased
evapotranspiration, the hydrograph will respond less
markedly to rainfall. Nevertheless, rainstorms may still occasionally
produce strongly peaked hydrographs. In winter,
river discharge is usually low or sometimes even absent.
Another characteristic of the periglacial fluvial
domain is the sometimes very high percent of the annual
runoff that occurs within the two to three weeks of the
snowmelt flood (break-up), with a corresponding concentration
of sediment transfer. Finally, the presence of permafrost
accentuates runoff peaks while reducing groundwater
supply to river runoff and thereby base flow.
in northern Russia were conducted (Dankers 2002, Van
der Linden 2002). The hydrographs of both river systems
fall within the category of the subarctic nival river regime.
The Tana River catchment is 16,000 km* in size; mean
annual discharge is only 166 m3/sec; interannual variation
is high; snowmelt runoff, with a maximum discharge as
high as 3,000 m3/sec, may contribute as much as 65 YO to
the total annual runoff; mean annual T varies from -0.5 to -
3°C and mean annual P from ca 350-450 mm; discontinuous
permafrost occurs. The Usa catchment is 93,000 km*
in size; peak discharges vary from ca 4,000 to more than
7,000 m3lsec; mean annual T varies from -3 to -7°C and
mean annual P from 400 -800 mm; continuous, discontinuous
and sporadic permafrost zones occur from north to
south.
The sensitivity of the discharge of the Tana and Usa
rivers to changes in climate was evaluated with hydrological
models, called TANAFLOW and USAFLOW, respectively,
in both cases based on a spatially-distributed water
balance model (RHINEFLOW) developed for the Rhine
River by Kwadijk (1993). Both models are embedded in a
GIS and calculate the water balance for grid cells of I by I
km; runoff is subsequently routed over a digital elevation
model (DEM) of the area and accumulates at the catch-
ment outlet. Originally designed for monthly time steps,
the models have now been applied on a 10-day basis.
Moreover, the USAFLOW model (Van der Linden 2002)
has been adapted for cold-climate conditions by introducing
variable (in space and time) separation coefficients -
dividing rapid (surface) runoff from water added to
groundwater storage - for continuous (1.0 100% rapid
runoff), discontinuous (0.9), sporadic permafrost (0.8) and
for seasonal frost (0.6).
WATER BALANCE MODELLING
In the framework of two EC-funded research projects,
“Barents Sea Impact Study” (BASIS) and “Tundra Degradation
in the Russian Arctic” (TUNDRA), hydrological
modelling studies of the Tana river in northern Fennoscandia
and the Usa River as part of the Pechora basin
IMPACT OF CLIMATE CHANGE
In general, it is expected that should warming occur, permafrost
will degrade, and storage capacity of the soil will
increase. Consequently, there will be fewer high flow
events of large magnitude, and the probability of extremely
low flows will also diminish. RIP ratios will de-
85

crease as groundwater storage increases, but surface flow
remains important during spring (Woo 2000, Van der Linden
2002). By introducing temperature and precipitation
changes concurrent with state-of-the-art climate scenarios,
the sensitivity of river discharge to these climate variables
was established (Harding et al. 2002). Fig. IA illustrates
the change in the Tana hydrograph between a
control run (simulated for the period 1980-1993) and a
scenario run (simulated for the period 2070-2100), derived
from the RCM HIRHAM4 model. The combined effects of
increased temperatures, precipitation and evaporation result
in a distinctly earlier timing (by about 20 days) of the
runoff peak, which is also slightly higher. The latter is due
to the somewhat larger snow mass, in spite of the shorter
snow season, and reduced sublimation. Although smaller
in amounts, the relative increase in runoff is even larger in
the rest of the year. As baseflow is higher, runoff in winter
has increased in the scenario run as well. Fig. IB illustrates
the change in the Usa hydrograph for the present
situation and according to a GCM HADCM2 scenario projection
to the period around 2080. It shows a decrease in
peak discharge. In this region the increased evaporation in
summer due to a rise in temperature is not compensated
by larger precipitation amounts, which causes discharge
to decrease. To further improve hydrological models, the
feedback effects of vegetation responses to climate
change should be taken into account. In conclusion, these
studies reveal that the indirect effects of climate change
(e.g., vegetation and permafrost changes) on discharge in
(sub)arctic areas are relatively small compared to direct
effects.
Koster, E. el al.: Hydrological modelling, climate change ...
REFERENCES
Dankers, R. 2002. Sub-arctic hydrology and climate change. A case
study of the Tana River Basin in Northern Fennoscandia. Netherlands
Geographical Studies 304,237 pp.
Harding, R., Kuhry, P., Christensen, T.R., Sykes, M.T., Dankers, R.
and Van der Linden, S. 2002. Climate feedbacks at the tundrataiga
interface. Ambio Special Report 12: 47-55
Kwadijk, J.C.J. 1993. The impact of climate change on the discharge
of the River Rhine. Netherlands Geographical Studies 168, 201
PP.
Van der Linden, S. 2002. Ice rivers heating up. Modelling hydrological
impacts of climate change in the (sub)arctic. Netherlands Geographical
Studies 308, 174 pp.
Woo, M-K. 2000. permafrost hydrology. in: Nuttall, m. and Callaghan,
T.V. (eds.): The Arctic, Environment, People, Policy. Harwood,
Amsterdam: 57-96
86

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Using BTS for mapping permafrost distribution in Apussuit, SW Greenland
L. Krisfensen
Institute of Geography, University of Copenhagen, Denmark
INTRODUCTION
Permafrost in Greenland has been mapped on a national
scale by Weidick (1968) and Christiansen and Humlum
(2000) but no previous1 attempts have been made to map
mountain permafrost on a local or regional scale. Geomorphological
indicators of permafrost such as pingos and
rock glaciers, as well as borehole temperatures from permafrost,
are reported from various locations in Greenland.
On the west coast of Greenland the southernmost location
of reported geomorphological indicators of permafrost is in
the Sisimiut area. Here, borehole temperatures show thin
permafrost at sea level.
The BTS Method (Haeberli 1973) has been succesfully
used for mapping mountain permafrost in the Alps and
other mountain regions but, so far, has not been used in
Greenland.
SITE DESCRIPTION
The Apussuit study area covers 30 * 40 km and is located
on mainland Greenland at 65”32’N, 52’22’E, 25 km NE of
Maniitsoq, approximately 30 km from the outer west coast
and 100 km west of the Inland Ice sheet. The area is 10-
cated in a zone of discontinuous permafrost and on the
border between sporadic and discontinous permafrost on
Weidick’s (1968) and Christiansen and Humlum’s (2000)
permafrost maps of Greenland. However, no previous
studies on permafrost have been carried out in the area.
The Apussuit ice cap sits on a narrow plateau at 800 -
1133 m a.s.1. and measures around 110 km2. Some of the
highest mountains of western Greenland are located north
of Apussuit, where large ice caps can also be found. The
mountains were probably nunataks during the Quartenary
glaciations (Weidick 1968). The landscape south of Apussuit
has a gentler, glacially-eroded topography. Sediments
are sparse due to low weathering rates of the archaic
gneisses, so apart from block fields and talus slopes,
geomorphological indicators of permafrost are rare. The
fieldwork was carried out near a small skisport center at a
nunatak in 931 m a.s.1. The mean annual air temperature
at sea level is around -1°C.
METHOD
A Digital Terrain Model (DEM) with a cell size of 40 * 40 rn
was interpolated from 150,000 point values originating
from photogrammetry of aerial photographs. The potential
direct solar radiation (PR) was calculated on an hourly basis
throughout the DEM and averaged over a year. The
calculations include damping of the radiation through the
atmosphere and shadow from the surrounding horizon.
160 BTS measurements at 81 points were recorded
between 925 m a.s.1. and sea level, together with GPS location,
snow depth, inclination and aspect. The measurements
were made at snow depths between 85 cm and 200
cm. The mean distance between the recorded GPS point
and the centre of the appropriate cell in the DEM was 15.1
m. The differences between measured and modelled inclination
and aspect were 8.4’ and 39.8” respectively. These
differences are most likely caused by differences in scale
as the parameters in the field were measured for an area
covering around IO* 10 m, whereas the cell size in the
model was 40 * 40 m.
RESULTS AND DISCUSSION
I30 measurements were in the “permafrost probable” BTS
class (BTS-3”C), 17 in the “permafrost possible” BTS
class (-3°C BTS -2°C) and 13 in the “permafrost improbable”
BTS class (BTS > -272, Hoelzle 1992). The
highest located point in “permafrost improbable’’ BTSclass
was at 283 m a.s.1.; the highest point in “permafrost
possible” BTS class was measured at 478 m, whereas several
points in the “permafrost probable” BTS class were
found near sea level. The results are supported by borehole
temperatures from the area showing permafrost at
87

Kristensen. 1,: Using BTS for mapping perniafrost distribution ...
360 m a.s.1.
The BTS measurements were analyzed with respect to
altitude, PR and snow depth. Within the measured interval
of snow depths, no significant relation between BTS and
snow depth was found. Altitude explains 59 % of the variance
in BTS and PR 25 YO. However, a significant negative
correlation (r -0,33) between PR and BTS shows
that the dataset is biased, indicating that areas facing
south are more likely to have been sampled at low altitudes.
Because of this multiple regression, using both altitude
and PR as the controlling variables only improves
R2 to 66 %. The relation takes the form:
BTS = -0.0054~ altitude + 0,028 x PR - 5.30
The equation was applied to the entire study area,
simulating BTS throughout the DEM. 121 BTS measurements
were assigned to the appropriate BTS class, 33
were assigned to the neighbouring BTS class, whereas six
were assigned to two BTS classes from the measured
one. According to this relation, permafrost limits are highly
influenced by radiation. Areas receiving low amounts of
radiation are modelled to belong to the “permafrost probable”
BTS class at sea level, whereas areas receiving
maximum radiation are modelled to the “permafrost improbable”
BTS class up to 250 m a.s.1. and “permafrost
possible” up to 430 m a.s.1. The importance of altitude and
PR in explaining BTS is comparable to findings from Jotunheimen
and Dovrefjell, southern Norway (approximately
62’), though permafrost in these areas is found at
higher altitudes (Odegard 1996). Radiation is probably of
much greater importance in the eastern Swiss Alps (approximately
46”, Hoelzle 1992) than in Greenland, which
might reflect the lower latitude and higher amounts of radiation
received in the Alps.
sults are supported by borehole temperatures from the
area showing permafrost at 360 m a.s.1. However, multiple
regression including both altitude and PR shows that the
thermal regime of the ground is highly influenced by PR as
permafrost-free areas or areas marginal to permafrost are
found up to nearly 500 m a.s.1. where PR is high.
REFERENCES
Christiansen, H. H. and Humlum, 0. 2000. Permafrost. In B. H.
Jakobsen. Bocher, J. Nielsen, N. Guttesen, R. Humlum, 0. and
Jensen, E (eds), Topografisk Altas Grernland. Kerbenhavn: C.A.
Reitzels Foriag.
Haeberli, W. 1973. Die Basis-Temperatur der winterlichen
Schneedecke als moglicher indicator fur die Verbreitung von
Permafrost in den Alpen. Zeitschrift fur Gletscherkunde und
Glazialgeologie 9: 221 -227.
Hoelzle, M. (1992). Permafrost occurrence from BTS measurements
and climatic parameters in the Eastern Swiss Alps. Permafrost
and Periglacial Processes 3: 143-147.
Weidick, A. 1968. Observations on some Holocene glacier fluctuatuins
in West Greenland. Meddelelser om Granland. 164(6): 1-
202.
0degArd. R. S. Hoel.de, M. Johansen, K. V. and Sollid, J. L. 1996.
Permafrost mapping and prospecting in southern Norway. Norsk
geogr. Tidsskr. 50: 41-53.
Figure 1. The effect of
found by the model.
PR on BTS classes and permafrost limits as
CONCLUSIONS
BTS measurements in the Apussuit region indicate widespread
permafrost in the area, even at sea level. The re-
88

Measuring rock glacier surface velocities with real time kinematics GPS
(Mont Gele area, western Swiss Alps)
C. lambiel, *R. Delaloye, **l. Baron and R. Monnet
Institute of Geography, University of Lausanne, Switzerland
*Dept. of Geosciences, Geography, University of Fribourg, Switzerland
**/nsfitufe of Geophysics, University of Lausanne, Switzerland
INTRODUCTION
Studies on rock glacier movement have been led for a
long time on several sites in the Alps (e.9. Barsch 1992).
Methods applied until now are essentially measurements
using theodolites (ens. Francou and Reynaud 1992) and
aerial photogrammetry (e.g. Kaab et al. 1997). Rock glacier
surface velocities measurements with Real Time
Kinematics (RTK) GPS have been tested in the Mont Gele
area (Verbier, Swiss Alps, 46"06'N/7"17'E).
The studies were conducted on three rock glaciers, two
of them (B and C) north-northeastern oriented and the
third one (D) facing towards the east. Rock glacier D is
mainly inactive (presence of depressed parts, no steep
front, abundance of lichens), whereas rock glaciers B and
C are clearly active (very steep fronts, unstable blocks, no
lichens). The roots of rock glacier C are largely covered by
an ice patch. The occurrence of permafrost is corroborated
in all rock glaciers by many BTS measurements
performed since 1993 and by continuous logging of the
subsurface temperature using miniature loggers UTL-1.
In summer 1998, DC resistivity soundings and mapping
lines were carried out on the rock glaciers (Reynard
et al. 1999). Some of the results are presented in Figure 1.
Marked differences in resistivities are primarily interpreted
as differences in ice content. They also could reveal differences
in temperature of the icehock mixture. According
to Reynard et al. (1999)) rock glaciers B and D contain
congelation ground ice and are therefore considered as
typical periglacial rock glaciers. Low resistivities (max. 50
kWm) should indicate a permafrost temperature close to
the thawing point. Feature C is more complex. The presence
of the ice patch at its roots and the high resistivities
measured (specific resistivity of about 350 kWm) indicate
the presence of massive (and colder) ice within the rock
glacier. The resistivity mapping demonstrated that there
was no break between the ice patch in the rooting zone
and the rock glacier (Reynard et al. 1999).
Figure 1. Three of the DC resistivity soundings carried out
Mont Gel6 rock glaciers (Reynard et al. 1999, modified).
METHOD
on the
The method used to survey the rock glacier surface velocities
is of Real Time Kinematics type (Leica Geosystems):
two stations (reference and rover) permits the local
coordinates to be localized after a few seconds with a
relative accuracy of 2-3 cm in the horizontal (x, y) coordinates
and 3-5 cm in the vertical coordinate (z).
In September 2000, 87 blocks were measured on the
rock glaciers. Four control points were also surveyed on
bedrock, i.e. at places that are not affected by movement.
Only blocks that seemed to be deeply buried in the active
layer (and if possible in the permafrost body) were chosen.
One year later, the position of the blocks was measured
again.
89

RESULTS AND INTERPRETATION
Results are presented in Figure 2. A spectacular difference
in horizontal movements between rock glaciers B
and C can be observed. Rock glacier B clearly shows the
maximal surface velocities. Annual displacement is up to
125 cm a-1 in the central part of the rock glacier. Such values
are relatively high compared to values usually obtained
on rock glaciers (Barsch 1992). The lowest movements
are encountered in the lateral parts of the
formation. The roots show smaller velocities than the central
and frontal parts of the rock glacier. Rock glacier C is
also affected by movements, but to a much lesser extent.
Maxima are encountered on the right side, toward the
frontal part. The left side and the upper flat area appear to
be more stable. Rock glacier D only shows velocities
lower than 5 cm a-I.
slope becomes steeper. This part corresponds to the
faster velocities. As no flat zone is present further down,
velocities remain high. Lower velocities measured on rock
glacier C, despite higher resistivities, can be explained by
topography as well. The slope of the formation is much
less regular and large flat areas dominate the rock glacier.
The presence of obstacles (rock glacier D and an outcrop
of roches moutonnees) at the front of the rock glacier can
also explain the relatively low velocities of this part. This
forces the rock glacier to turn with an angle of about 90"
to the right. Such a change in direction probably affects
the velocity of the rock glacier.
The second hypothesis refers to differences in icelrock
mixture temperature. Rock glaciers with temperature close
to the thawing point would creep faster than colder rock
glaciers (e.g., lkeda et al. 2003). Nevertheless, except for
the difference in electrical resistivity, there is still no further
data available confirming any difference in the temperatures
of both rock glaciers B and C.
CONCLUSION
RTK GPS has proved to be very efficient for rock glacier
surface velocity measurements. The precision of a few
centimeters is sufficient and one (long) day of work permits
more than a hundred points to be measured.
This ongoing study has suggested that surface velocities
of rock glaciers are not governed by ice content, but
are more closely related to the topography (slope angle) of
the rock glacier or to the temperature of the icelrock mixture.
REFERENCES
Horizontal surface velocities (cm a")
A 100to 125 F 11 to25 0

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Detection of frost heave and thaw settlement in tundra
environments: application of differential GPS technology
J. Little, H. Sandall, M. Walegur and F. Nelson
Department of Geography and Center for Climatic Research, University of Delaware, Newark, USA
Geocryologists have devised several methods to measure
the vertical component of frost heave and thaw subsidence.
Most have employed either (a) mechanical
“heavometers” that record differential movement at points
or over very small (e.g. 10 m*) areas; or (b) optical leveling,
used in conjunction with specially designed platform
targets. Measurements are usually made within localized
coordinate systems and have rarely been related precisely
to geodetic coordinate systems.
Degradation of permafrost and thaw settlement have
received much attention recently in popular and globalchange
literature (e.g. McCarthy et al. 2001, Goldman
2002). These processes pose considerable hazard to human
activities and infrastructure in high latitude regions.
Recent technological advances in geodetic science have
the potential to measure vertical movements accurately
and rapidly, to survey relatively large areas, and to communicate
survey results in precise, externally referenced
coordinate systems. Among the most promising technological
developments are Differential Global Positioning
Systems (DGPS).
During summer 2001 and 2002, post-processed sfopand-go
kinematic, rapid-static carrier phase DGPS, and a
total of sixty-four specially designed heave and subsidence
targets were used to record frost heave and ground
subsidence at two 1 ha sites in northern Alaska. One site
is at Sagwon in the northern foothills of the Brooks Range,
the other at West Dock in the Prudhoe Bay oilfield on the
Coastal Plain (Fig. 1). Our preliminary results indicate that
post-processed rapid-static carrier phase DGPS has the
ability to measure frost heave and ground subsidence at
very high resolution (e.g. I cm).
Global Positioning Systems acquire coordinate positions
through triangulation of an antenna receiver with at
least four satellites. DGPS, which requires two antenna
receivers, improves the accuracy to the cm scale; we used
a base receiver (Trimble 4000) and a rover receiver
(Trimble 4700). The inexpensive platform targets are cylindrical
platforms capable of supporting a rover DGPS
antenna while allowing the freedom to move vertically in
response to heave or settlement. The ability to move the
antenna from target to target permits acquisition of data at
time scales varying from 15 to 45 seconds for stop-and-go
kinematic DGPS; rapid-static DGPS requires eight to
twenty minutes, but improves accuracy substantially (Little
et al. in review).
The West Dock and Sagwon sites, located in the Kuparuk
River basin of northern Alaska (Fig. I), were instrumented
with 32 platform targets, arranged according to a
nested hierarchical sampling design (Nelson et al. 1999).
The West Dock site is composed primarily of wet nonacidic
tundra, and is underlain by ice-wedge polygons.
The study site occupies part of a drained thaw lake basin
and an adjacent section of “upland” tundra, which has
better drainage. The Sagwon site occupies a section of
nonacidic tundra with frost boils and patchy vegetative
cover (Walker et al. 1998).
Figure 1. Map showing locations of West Dock and Sagwon (Flux 3)
sites, North Slope, Alaska near the Dalton Highway.
All targets were given an arbitrary position of 10 cm at
the beginning of the survey during summer 2001; Figure 2
shows heave and subsidence relative to this position.
Rapid-static DGPS observations for June 2002 resulted in
average surface heave of 1 cm from July 2001 (Fig. 2a)
with maximum heave and subsidence of 6.7 cm and 2 cm,
respectively. Although observations were recorded in July
2001 instead of August, the normal period of maximum
thaw in northern Alaska, 24 of 32 targets recorded heave
by June 2002. Mean values from a rapid-static DGPS survey
in August 2002 were 4.3 cm lower than in June 2002.
91

Subsidence was recorded at 29 of 32 targets, with maximum
subsidence of 15.5 cm and heave of 0.5 cm (Fig.
2a).
Heave of 1.9 cm was recorded from August 2001 to
June 2002. Of the 32 targets, 24 recorded heave in June
2002 (Fig. 2b). Rapid-static DGPS from August 2002 recorded
3.3 cm of subsidence from June 2002 (Fig. 2b).
2a 20
E
0 T'2L
-5 n
-10
June 02 August 02
2b 30 7
2c
-In
T
June 02 August 02
40 3 T
classical surveying and rapid-static DGPS, respectively,
yet were 14.2 for stop and go kinematic DGPS (Fig. 2c).
The close correspondence between traditional survey
methods and rapid-static DGPS at West Dock indicates
that frost heave and thaw can be successfully detected
with the latter. As part of an ongoing project, the vertical
displacement of each platform will be measured again
using DGPS in summer 2003.
ACKNOWLEDGEMENTS
This DaDer is a contribution to th e CircumDolar Activ e
Laye; Monitoring (CALM) program. We are grateful to
Shad O'Neel (UNAVCO) for making DGPS equipment
available for work in Alaska and for formal instruction in its
use at UNAVCO's Boulder, CO facility. Professor Carmine
Balascio (Department of Civil Engineering, University of
Delaware) kindly provided surveying equipment for the
validation work. We thank British Petroleum (Alaska) Inc.
for logistical support and its Health, Safety and Environment
office for access to the West Dock site. Funding was
provided by the US. National Science Foundation under
Grant OPP-0095088 to F.E. Nelson and Grant OPP-
9732051 to K.M. Hinkel (University of Cincinnati). Any
opinions, findings, conclusions, or recommendations expressed
in this material are those of the authors and do
not necessarily reflect the views of the National Science
Foundation. Identification of specific products and manufacturers
in the text does not imply endorsement by the
National Science Foundation.
Classical Rapid Static Stop and go kinematic
Figure 2a. Box plot depicting heave June 2002 from July 2001 at
West Dock. Subsidence occurred between June and August 2002 and
was detected by rapid-static DGPS. Figure 2b. Box plot showing
heave and subsidence at Flux 3. Figure 2c. Box plot comparing results
from traditional surveying with stop and go kinematic and rapidstatic
DGPS.
The use of DGPS as a stand-alone replacement for
traditional survey methods (e.g., differential leveling or
profile leveling) has been called into question (e.g.! US
Army Corps of Engineers 1996). To ascertain the degree
of correspondence between DGPS and traditional survey
techniques in the context of frost heave and thaw settlement,
we compared results at West Dock from rapid-static
and stop and go kinematic DGPS with those from a survey
using profile leveling. The traditional survey, conducted in
August 2002, detected mean subsidence of 2 cm, compared
with results from June 2002 (Fig. 2c). Comparison
of results from traditional surveying and the two DGPS
techniques confirmed that rapid-static DGPS yields much
better accuracy. Profile leveling in June 2002 averaged
just 1.2 cm higher than rapid-static DEPS recorded a few
days earlier, but 5 cm higher than stop-and-go kinematic
(Fig. 2c). Standard deviations were 1.6 and 3.9 cm for
REFERENCES
Goldman, E. 2002. Even in the High Arctic, Nothing is Permanent.
Science, 297: 1493-1494.
Little, J.D., Sandall H., Walegur, M.T., and Nelson, F.E in review. Application
of differential GPS to monitor frost heave and ground
subsidence in tundra environments. Permafrost and Periglacial
Processes.
McCarthy, J.J., Canziani, O.F., Leary, N.A., Dokken, D.J. and White,
K.S. 2001. Climate Change 2001: Impacts, Adaptation, and Vulnerability.
Cambridge University Press, Cambridge, UK.
US Army Corps of Engineers. 1996. Engineering and Design -
NAVSTAR Global Positioning System Surveying. April, 2002.
http:I/www.usace.army.miI/usace-docs/eng-manuals/em111O-1-
1003/toc.htm.
Walker, D.A., Auerbach, N., Bockheim, J.G., Chapin, F.S. 1.1.1,, Eugster,
W., King, J.Y., McFadden, J.P., Michaelson, G.J., Nelson,
F.E., Oechel, W.C., Ping, C.L., Reeburgh, W.S., Regli, S., Shiklomanov,
N.I., and Vourlitis, G.L., 1998: A major arctic soil pH
boundary: implications for energy and trace-gas fluxes. Nature
394: 469-472.
92

Permafrost and Little Ice Age glaciers relationships: a case study in the
Posets Massif, Central Pyrenees, Spain
R. lugon, *R. Delaloye, **E. Serrano, ***E. Reynard and C. lambiel
University Institute Kurt Bosch, Sion, Switzerland
*Department of Geosciences, University of Fribourg, Switzerland
**Department of Geography, University of Valladolid, Spain
***Institute of Geography, University of Lausanne, Switzerland
INTRODUCTION
The prospecting of ground ice in sediment deposits within a
glacier forefield requires two main types of ground ice to be
differentiated: pure permafrost ice (congelation ice within the
ground) and debris-covered (or buried) glacier ice. DC (direct
current) electrical prospecting is an easy and light technique,
which allows such a distinction to be revealed in many situations
(Haeberli and Vonder Muhlll996).
Resultsfromageoelectricalmeasuringcampaignin the
forefield of the Posets glacier are briefly presented and discussed
here. As far as we know, this is the first study treating
the relationship glacier - permafrost (rock glacier) within
the Little Ice Age (LIA) Pyrenean glacier forefields.
Hummel configuration allowed detection of shifts in ground resistivity
in both branches of the soundings.
The VES were partially complemented with resistivity
mapping (wenner configuration). The technique allowed spatial
changes in the measured apparent resistivity (usually 2-
10 times lower than the specific resistivity for a permafrost
body) to be determined at a fixed pseudo-depth.
A challenging aspect was to dispose of a sufficient electrical
supply for a whole week of measuring. An OYO
McOhm device (model 2115), complemented with a single
lithium-lead battery (about 10 kg), was well suited as no recharge
was required. This was only possible due to the use
of low-intensity current (between 1 and 5 mA). Sponges
soaked with salt water were used as electrodes.
STUDY AREA
The LIA margins of the Posets glacier are located between
2900 and 3100 m a.s.1. on the eastern side of the Posets
peak (3369 m a.s.1.) in the Central Pyrenees (42"39'N,
0'26'E). In 2001, the Posets glacier covered less than 0.1
km2 (Fig. I). It was three to four times larger during the LIA.
At that time, the glacier divided into two tongues, one flowing
to the east towards the Posets rock glacier, and the other b
the north towards the La Paul rock glacier. The presence d
permafrost in the glacierhock glaciers complex was shown
by Serranoet ai. (2001) who used a setof BTS (Bottom
Temperature of the Snow cover) and year-round continuous
ground temperature measurements.
The bedrock beneath almost all the glacial sediment aG
cumulations consists of a granite batholith. Hornfels resulting
from contact metamorphism of Paleozoic shales surrounds
the igneous intrusion and forms the greater part of the face an
Posets peak overhanging the glacier cirque.
PROSPECTING METHODS
Two complementary DC resistivity techniques were applied.
Vertical electrical soundings (VES) were used to estimate the
stratigraphy of the ground resistivity. The dissymmetrical
RESULTS AND INTERPRETATION
Posefs rock glacier
Two VES were performed on the Posets rock glacier, a 400
m long imposing sedimentary formation located at the front d
the former eastern tongue of the Posets glacier (Fig.1). Five
main arched ridges are clearly visible. Under a thin (1 m) active
layer (20 kam), mainly consisting of angular homfel
clasts, an extreme resistive layer was detected (specific re
sistivity: 3-12 MQm, 5-12 m thick), beneath which the resistivity
falls towards few a tens of kQm.
The spatial extension of the apparent resistivity of the
second layer was mapped at a depth of about 6-9 m (AB
37.5 m, 78 measurements). Extreme apparent resistivities,
somewhere higher than 500 kQm, were measured in the
whole sedimentary complex. However, resistivities strongly
decreased towards the rooting zone of the debris rock glacier
where values between 50 and I00 kQm were found.
The high resistivity of the second layer clearly indicates
the presence of a massive ice layer, immediately beneath a
thin layer of deposits. Such massive ice is interpreted as
buried ice of glacial origin. The very high resistivity of the
second layer could have madedeeper layer(s) with lower
resistivities undetectable, i.e., permafrost with low resistive
ground ice.
93

la Paul rock glacier
The 400 mlongLa Paul rock glacier shows a typical but
poorly-developed relief of permahst creep with transversal
furrows and ridges. A VES revealed the presence of an extremely
resistive layer (8-25 MQm, 5-15 m thick) below a 2
m thick active layer (3040 kQm). The downward branch d
the VES shows the decreased resistivity of the second layer
towards the frontal part of the rock glacier (1.3-5 MQm, 515
m thick). As for the Posets rock glacier, geoelectrical methods
could not prove nor exclude the existence of a third underlying
frozen layer.
The extreme values of resistivity calculated towards the
rooting zones of the rock glacier are only known to be found
in cold or polythermal glaciers (buried sedimentary ice, Haeberli
and Vonder Muhlll996).
FURTHER QUESTIONS
Geo electrical measurements pointed to extreme resistivity
ice in two debris rock glaciers that were prospected. Specific
resitivities ranging between 1 and 25 MQm were detected
under a thin (1-2 m) active layer. The following suppositions
can be made about the origin of ice within both Posets La and
Paul rock glaciers.
lugon, R. et 81.: Permafrost and Little Ice Age glacicrs ...
Debris-covered glacier ice could have been incorporated
in the two rock glaciers during the advance of the Posets glacier
in the LIA, or former Holocene advances. Very cold
subsurface temperatures measured in the deposits (Serrano
et al. 2001) indicate that buried ice could easily have been
preserved since the end of the LIA.
The massive ice probablysuperimposedformerpermafrost
bodies, i.e., a much thicker layer of perennially frozen
ground. However, such permafrost could not be detected by
means of geo-electrical techniques.
The processes which have created both La Paul and
Posets debris rock glaciers therefore remain in question.
What is the role played by the glacier dynamic in the
formation of the deposits What is the importance of the
creeping processes Do the ridges occurring at the surface d
the Posets glacier attest to advances of the Posets glacier
during LIA, or do they indicate a sign of permafrost creeping
REFERENCES
Haeberli, W. and Vonder Muhll, D. 1996. On the characteristics and
possible origins of ice in rock glacier permafrost. Zeitschrift fur
Geomorphologie, Suppl. Bd. 104: 43-57.
Serrano, E., Agudo, C., Delaloye, R. and Gonzalez-Trueba, J.J.
2001. Permafrost distribution in the Posets massif, Central
Pyrenees, Norsk Geografisk TidsskrifbNorwegian Journal of
Geography 55: 245-252.
Little Ice Age Pnsets ~ l ~ ~ i e ~
94

Borehole drilling in permafrost of an avalanche-affected slope at
Fluelapass, eastern Swiss Alps
M. luetschg, V. Sfoecki, W. Ammann and *W. Haeberli
Swiss Federal institute for Snow and Avalanche Research (SLF), Davos Doe Switzerland
*Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Zurich, Switzerland
Alpine permafrost is affected by different topographic and
climatic factors. The snow cover is of high importance due
to its strong insulation and albedo effects. Snow cover has
a warming effect on ground temperatures during cold
winter, whereas in late spring and summer, snow cover
has a cooling effect (Keller 1994). This means that not
only altitude and aspect but also snow-cover distribution is
important with regard to the permafrost distribution; avalanches,
therefore, play an important role in permafrost
distribution, as they redistribute, compress and accumulate
the snow: hence, at the accumulation zone of an
avalanche slope, cooler ground temperatures were measured
than in the passing zone higher above, due to the
long remaining avalanche snow (Haeberli 1975). Thus,
permafrost often occurs in the lower parts of avalanche
slopes.
Such an avalanche-affected permafrost slope was
found at Fluelapass, located in the eastern Swiss Alps, at
an altitude of 2380 m a. s. I. The slope above the lake
Schottensee is exposed towards northeast and is between
33 and 37 degrees steep at the upper part of the slope,
corresponding to a typical avalanche starting zone.
Avalanches indeed occur in this slope in late winter
and spring and deposit the snow at the foot of the slope.
These large snow accumulations disappear later in the
season than the snow on the slope above. This fact was
also observed by a series of automatically-taken photographs
during I999 (Lerjen et al. 2003). Measurements of
ground surface temperatures and seismic refraction
soundings indicate colder temperatures and the presence
of ice close to the surface in the lower third of the slope
(Haeberli 1975). In summer 1999, these measurements
were repeated and the active layer thickness was determined
to be between 2 and 4 m in the lower part of the
slope (Lerjen et al. 2003).
Ground surface temperatures in the slope and at both
shores of the lake were measured during winter 2001102.
In summer 2002, two boreholes were drilled, one (bh I)
in the avalanche starting zone (7913521180353 2500 m a.
s. I.) and one (bh 2) at the foot of the slope
(7915001180474 2394 m a. s. I.) to study the effect from
the redistribution of snow due to avalanche events on
ground temperatures. In the lower borehole (bh 2), cores
containing pore ice were sporadically drilled. Figure l-A
shows the stratigraphy of the two boreholes as based on
information from the experienced borehole-drilling team,
and Figure l-B presents the temperature profiles which
were measured at the beginning of winter 2002103, six
weeks after the end of the drilling.
In the upper borehole (bh I), 40 cm of humus to sandy
soil was found under a dense vegetation cover (Fig. 1-A).
This top layer was underlain by blocky material which
contained larger blocks in the lower part. At 14.5 m depth,
bedrock was encountered. The lower borehole (bh 2) differs
from the upper borehole by the lack of a vegetation
cover and the presence of pore ice: this ice was found
between 1.8 and 8.5 m depth. Cores were drilled at the
depths from 3.3 to 4.25 m and from 5.1 to 6.44 m; they
contained some fragments of the pore ice. In the lower
borehole, bedrock was not reached but at 20 m depth, the
porous material was filled with water, indicating that the
level of the lake surface was reached.
In both boreholes, inclinometer tubes were inserted to
measure future deformation of the scree slope. Additional
surveyings of the borehole positions (surface coordinates)
are performed twice a year. Ground temperatures have
been monitored continuously since October 2002.
First borehole temperature measurements at the beginning
of winter 2002103 reveal a clear difference between
the two positions in the slope. In the lower borehole
(bh 2), temperatures are close to 0°C (fig. 1-B), while
below 13 m, the temperature in borehole bh 2 increases to
attain positive values. Note that the spatial resolution of
the temperature profile in the lower part is 5 m, corresponding
to the distance of the installed temperature sensors.
Since winter 2001102, surface temperatures at four positions
between the two borehole positions were measured
continuously with universal temperature loggers
(UTL) (Fig. 2). They show that the UTLs placed at lower
altitudes close to the lower borehole position are becoming
snow-free about 80 days later than in the upper part of
95

the slope. BTS were measured at the end of April 2002 to
be - 2.8 "C in the upper part and - 3.8 "C in the lower
part of the slope.
The borehole temperatures are supposed to still be
somewhat affected by the effects of drilling disturbance,
due to the short period between the end of drilling and the
first temperature measurements presented in Figure 1-B.
Results from the studies presented here will be used in
comparison to modelling results, in which the situation is
calculated with the one-dimensional model SNOWPACK
(Bartelt and Lehning 2002). This snow-cover model was
developed and extended to the underlying ground. It will
be used to study the effect of snow height and snow density
on ground temperatures in the avalanche starting
zone and at the foot of the slope.
Borehole temperature measurements, surveys of snow
cover characteristics and snow distribution and numerical
simulations will give a better understanding of the effect of
avalanches on permafrost distribution in avalanche
slopes.
T["CI
bh 1 bh 2 P l ! i l I
ACKNOWLEDGEMENTS
The borehole drilling presented here was carried out by
Stump Bohr AG. This work was partly funded by the Swiss
National Science Foundation (Nr. 21-65263.01).
REFERENCES
Haeberli, W. 1975. Untersuchungen zur Verbreitung von Permafrost
zwischen Fluelapass und Pit Grialetsch (Graubunden). Mitteilung
der Versuchsanstalt fur Wasserbau, Hydrologie und Glaziologie
der ETH Zurich 17: 221.
Keller, F. 1994. lnteraktion zwischen Schnee und Permafrost. Mitteilung
der Versuchsanstalt fur Wasserbau, Hydrologie und Glatiologie
der ETH Zurich 127: 145.
Bartelt, P. and Lehning, M. 2002. A physical SNOWPACK model for
the Swiss avalanche warning. Part I: numerical model. Cold Regions
Science and Technology 35(3): 123-145.
Lerjen, M., Kaab, A., Hoelzle, M. and Haeberli, W. 2003. Local distribution
pattern of discontinuous mountain permafrost. A process
study at Fluela Pass, Swiss Alps. 8th International Conference on
Permafrost, Zurich, Switzerland.
blocks, stoner,
Ice
blocks. stones
block
stnnm
mo-e
block
sohd rock
blocks.
stoncs.
blocks.
stoner.
porous
blocks
mome
block. stones,
WatB
I
I
I
I
A
B
Figure 1. Stratigraphy of the the upper borehole (bh 1) and the lower
borehole (bh 2) drilled 2002 in an avalanche slope at Fluelapass (A),
and the corresponding ground-temperature profiles T with depth z,
measured six weeks after drilling (B).
30 , I
Figure 2. Ground surface temperatures (Ts) measured in 2001/02
between the two borehole positions in the avalanche slope at Fluelapass
show a temperature inversion with altitude.
96

Research on the relationship between wave velocity and the drillability of
frozen clay
Q. Y. Ma, *W. Ma, **M. F. Cai and ***Z. H. Zhang
Anhui University of Science and Technology, Huainan Anhui 232001, China
*University of Science and Technology of Bejing, Beijng 700083, China
"Sfate Key Laboratory of Frozen Soil Engineering, Cold and Arid Regions Environmental and Engineering Researching
Institute, CAS, Lanzhou 730000, China
**University of Science and Technology of Bejing, Beijng 100083, China
***Anhui Univemty of Science and Technology, Huainan Anhui 232001, China
From studies on the theory of elasticity we know that the wave. A specially made constant temperature device for b
propagation of an acoustic wave in rock or frozen ground is
closely related to the character of the medium. Once the lon-
Zen clay and an axial pressure coupled measuring shelf
were used to do the experiment. Measurements were made
gitudinal wave velocity and the transverse (shear) wave to a precision of kO.1 "C at six temperatures of: -5"C, -7"C,
velocity have been determined, the corresponding character- -lO°C,-12"C,
-15Co,,and -17°C respectively. The experiistics
of elastic deformation are also defined. The propagation mental results are shown in Table 1.
velocity of the acoustic wave, as measured in frozen ground,
reflects the likely drilling properties of this frozen ground. The
propagation velocity of frozen ground represents some typical
Table 1, The experimental results of longitudinal wave
velocity and transverse wave velocity of frozen clays
physical properties. Using this basic principle the relationship
name temperature longitudinal wave transverse wave
between wave velocity and the drillability of frozen ground
"C velocity, mls velocity, mls
was derived. The measurement of wave velocity has certain clay C1 -5 2363 1001
advantageous characteristics. This test is quick to perform clay ~2 -7 2477 1077
and not intrusive, hence it causes no disturbance to the clay C3 -10 2594 1170
ground. Using the wave velocity to determine the drillability clay C4 -12 2653 1224
of frozen ground is of great practical value.
clay C5 -15 2725 1279
clay C6 -17 2749 1300
A segment of frozen clay at a depth of 219.93-224.82m in
a mine shaft in the Anhui province of China was chosen. Tne Measuring the drillability of frozen clay requires that the
density of clay was 17lOkglm3 and had a water content d crushing instrument or chisel should follow an identical course
21.42%.
to the actual drilling hole so that the measured indexes of the
In the experiment, a UVM-2 acoustic wave instrument drillability do not vary with time, position or person, this leads
was used. This instrument can, via the sing-around method, to an objective and scientific index. Besides, using the chisperform
a delayed determination so that it is not affected by eling work per unit volume to reflect the drillability of frozen
multiple reflection waves and the velocity of the acoustic clay makes it very comparable.Theimpact is easily conwave
in the material can be measured rapidly and accu- trolled and sufficient load may be applied, even from a portrately.
The method used is as follows: the sensor is covered able instrument, to crush frozen clay of any typical hardness.
with petroleum grease and placed on both sides of the sam- Refemng to the research on the available experimental
ple. The receiver position is adjusted while observing the sig- equipment for drilling in rock, a chiselling instrument made by
nal through an oscilloscope. The location and width of the os- East North University has been chosen.
cilloscope display should be adjusted so that the whole The experimental method is as follows.
ultrasonic pulse can be seen in the window and an optimal
(I) The moisture content and particle size distribution should
wave form is attained. The measured propagation time is the be measured before the experiment.
measured acoustic cycle minus both the sensor delay and (2) The sample should be fixed to the test frame within the
the fixed delay. The acoustic velocity can be calculated on low-temperature room.
this basis.
(3) A disc, graduated into 24 equal parts, with a central
The diameter of the experimental sample was 61mm, the hole of 60mm diameter was fixed to the sample.
height of both the longitudinal and transverse wave samples
was 50.7mm. The experimental system had a fixed delay d
2.99s for the longitudinal wave and 2.00s for the transverse

tus. One person needs to ensure that the angle setting
knob does not move while the other operates the hammer.
It must be ensured that the hammer can drop
freely without any restriction.
(5) The angle setting knob was tumed 15" after each impact.
The pieces of frozen clay that fell to the bottom d
the hole were cleaned out after each five revolutions. In
addition, the hole depth was measured after 240 im-
pacts (IO revolutions) and 480 impacts (20 revolutions).
(6) The hole depth was measured with special vernier calipers
on the left edge, in the middle, and on the tight
edge of the hole, with a precision of 0.1 mm. Assuming
the frozen clay surface was smooth, the measured average
value was regarded as the hole depth.
(7)The chiseling work per unit volume can be calculated
with the chiseling depth as follows:
A
ne"=-
V
Ma! Q.Y, el al.: Research on the relationship between wave velocity ...
40NA,
d2H
Where a is the chiseling work per unit volume, kgmlcm3;
is the number of cycles, its actual value is 480; AO is the
work from each impact with an actual value of 3.9kg.m; d is
the diameter of the drilled hole, the actual diameter is lmm
greater than that of the drill bit so, value d is 4.1 cm; H is the
depth of the drilled hole in rnm and V is the volume in cm3.
The experimental results are shown in Table 2.
Table 2. Drilling depth and chiseling work per unit volume of frozen
Ma, Q. Y., Peng, W. W. and fhu, Y. L. 2002. Relationship
clays
Temperature, drilling depth, chiseling work per unit volume, artificially frozen clay. Chinese Journal
c",, mm kgo@/cm3
Engineering 21 (2): 290-294.
-5 122.20 11.90
Zhang, Z. H. and Ma, Q. Y. 2001. Research present situation
-7 116.57 12.48
analysis on classification of rock drillability, Chinese
-10 86.27 16.86
Coal Science and Engineering 7(1): 39-45.
-12 74.27 19.58
-15 67.17 21.65
-17 57.50 25.30
It can be seen from the data in Table 2 that the longitudinal
wave velocity, the transverse wave velocity and the chiseling
work per unit volume of frozen clay all appear to display
a tendency to increase as the temperature decreases. The
equation relating the longitudinal wave velocity, the transverse
wave velocity and the chiseling work per unit volume
can be established according to the principle of regression
analysis.
The regression equations are expressed as follows:
a= 0.807"~104~p2- 0.379cp -t 457.66, R~0.988
Where: cp= longitudinal wave velocity, mls.
a= 1.171"~10~~*~0.227~~ + 121.33, R=0.990
Where: cs= transverse wave velocity, mls
The multi-variable regression equation for longitudinal and
transverse wave velocity and chiseling work per unit volume
is expressed as follows:
a = -0.158~~ + 0.244~~ + 139.849, R=0.973, F= 54.507
The value of Fis 54.507 and can be derived. It is known
that F0.01(2,3)=30.8, F0.005(2,3)=49.8 and F0.001(2,3)=148.5
from the f distribution table. From this we get f > f0.005(2,3),
which signifies that the regression effect is excellent and that
the result of the multi-variable linear regression is acceptable.
N
From the results of the experiment we know that the longitudinal
and transverse wave velocity and the chiseling
work per unit volume of frozen clay all tend to increase as the
temperature decreases. While the longitudinal and transverse
wave velocity is higher and the frozen clay is harder, the
drilling is more difficult; conversely, when the longitudinal and
transverse wave velocity is lower, the drilling is relatively
easy. The relevant equation between the longitudinal and
transverse wave velocity and the drillability of frozen clay
can be established fairly easily by adopting the method d
multi-variable regression analysis. It indicates that using the
longitudinal and transverse wave velocity to evaluate the
drillability of frozen clay is feasible.
98
ACKNOWLEDGEMENTS
The authors wish to express their gratitude to The Outstanding
Youth Foundation for Scientific and Technological Research d
Anhui Province (2001-28), State Key LaboratoryofFrozen
Soil Engineering and Education Department of Anhui Province
of China for their financial support.
REFERENCES
Peng, W. W. and Yuan, Y. Q. 2000. Developing of a broadband ultrasonic
transducer of frozen soils. Journal of Glaciology and
Geocryology 22 (I): 85-89.
among
longitudinal and transverse wave velocities and temperature of
of Rock Mechanics and
and
journal of

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available i ~ f
Permafrost active layer as the concentrator of contaminants in northern
regions
V. N. Makarov
Permafrost lnsfifufe SB RAS, Yakufsk, Russia
Soils are said to be the “self-purification filters” of nature.
However, they lose their decontaminating properties in
permafrost zones. The reasons are: thinness of the soil
profile, inadequate drainage, thermohydrogeochemical
barrier to the migration of contaminants, annual freezing
and consequent concentration of soil water contaminants,
low geochemical and lowered chemical activity of soils
(because of low temperatures), short biological life of soils
during the year. These conditions cause increasing soil
pollution under the pressure from industrial zones.
Regional permafrost aquicludes at shallow depth from
the ground surface and low negative temperatures cause
a specific geochemical situation in the northern territories
and a widespread development of cryogenic metamorphoses
in underground waters. The common direction of
this process is to rise mineralization; the amounts of chloride
and sulfate increase in the waters and heavy metals
accumulate. During repeated freeze-thaw cycles the majority
of dissolved contaminants become concentrated in
gravitational waters while forming cryopegs. Cryopeg formation
is due to the intensity and duration of industrial activity
and results in its localization in the older parts of
populated areas (Makarov, Venzke 2000).
Studies of soil solution behavior indicate their functional
peculiarities during annual cycles. The most contrasting
changes of macro- and microelement fluxes, pH value
and oxidation-reduction potential (these determine migration
in the geochemical environments of the upper horizons
of seasonally thawed layers) are observed before
and after the cold period when sharp temperature changes
and moisture phase transitions occur. Before the beginning
of the cold period, the upper horizons of seasonally
thawed layers accumulate most of the components
that come from industrial geochemical fields. Cryogenic
concentration and the migration of soluble contaminants
deep into seasonally thawed layers, induced by freezing
front action, results in a depletion of the high horizons in
seasonally thawed layers during winter. It causes a reduction
in the concentrations of contaminant components by
the beginning of the warm period. Maximum seasonal fluctuations
in the intensity of geochemical fields is typical for
components such as Cr, Ag, Pb, Mo, S, Ca, Na, CI, Mg,
Sn whose concentration in pore water changes by 3 to 30
times (Fig. 1).
f” /
Figure 1, Seasonal variations in the concentration of chemical components
in soil’s pore water in Yakutsk
The penetration of anomalous industrial components to
the top of permafrost depends on the concentration gradient
and the geochemical properties of the chemical elements
forming the anomaly.
Because of the similarity in some chemical properties
and ion radii, iron and manganese hydroxides can sorb
heavy metals from a suspension and co-precipitate with
other metals (Ponnamperuma 1969). Heavy metal retention
by iron and manganese hydroxides increases with
time due to metal diffusion into the solid phase. During
periods of thawing soils become hydromorphic. This promotes
the formation of a large number of amorphous iron
hydroxides which increases the maximum adsorption of
heavy metals (Ostroumov 1998). However, the presence
of organic matter, even in small amounts, retards the
crystallization of iron hydroxides, thus causing increased
metal mobility. The highly mobile components of organic
matter in frost-affected soils form stable organometallic
components which have a high ability to migrate. Metal
mobility in organic soils is strongly influenced by microbial
metabolism. Thus, metal mobility in frost-affected organic

tion of three soil components, iron components, organic
matter and microbial communities.
The complicated combination of simultaneously occurring
processes determines the content ratio of elements in
the solid and liquid phases of industrial geochemical
fields. We can summarize the effects of the processes
using the thickness ratio of the chemical elements aureole
zone in the solid and liquid phases. Absorption of the solid
phase is predominant for some elements and dissolution
in water predominant for others.
The majority of contaminant solid phases is accumulated
in the upper parts of the ground at depths of 15-20
cm. The depth of contaminant penetration can reach 2.5-
2.7 m, including the top of the permafrost, depending on
the source of industrial contaminants and the intensity of
the cryogenesis processes (Makarov 1986). It is caused
by the transition of components in soil solutions and
ground waters and by further sedimentation on various
sorbents (Fig. 2).
Makarw, V.N.: Permafrost active layer as the concentrator ...
Figure 2. Distribution of chemical components in the active layer and
permafrost. 1 - anthropogenically affected soils; 2 - sand; 3 - sandy
silt; 4 - clay; 5 - permafrost.
The thickest and ecologically
most dangerous con-
taminant anomalies - Hg, Pb, P, s, Ag, Sn, Cu, Y, Zn, Cr
are forming at the thermohydrogeochemical barrier in urban
zones.
REFERENCES
Makarov, V.N. 1986. Geochemistry of frost-affected soils in urban areas.
In: Geochemistry of Technogenesis, Novosibirsk, Nauka:
118-124.
Makarov, V.N., Venzke, J.-F. 2000. Umweltbelastung und Permafrost
in den Stadt- und Agrarokosystemen in Zentral-Jakutien, Sibirien.
Geographische Rundschau 12, Braunschweig: 21-26.
Ostroumov, V.E. 1998. Ion redistribution in soils upon freezing.
Pochvovedenie 5: 614-619.
Ponnamperuma F. N., Loy, T. A. and Tianco, E. M. 1969. Redox
equilibria in flooded soils: The Manganese oxide systems. Soil
Science, 108, 1 : 48-57.
100

Borehole and active-layer monitoring in the northen Tien Shan
(Kazakhstan)
S. S. Marchenko
Permafrost Institute, Russian Academy of Sciences, Kazakhstan Alpine Geocryological laboratory, Almaty, Kazakhstan
The Zhusalykezen Mountain pass (43"05'N, 76"55'E,
3330 m a.s.1.) is located within the discontinuous permafrost
area of the Transili Alatau Range (frontal range of
northern Tien Shan). The location is close to the city of
Almaty, and therefore offers an opportunity for geocryological
and ecological investigations of the natural system
and systems affected by human activities.
The Upper Pleistocene and Holocene moraines from
the mountain descend into the nearby valleys. Vegetation
cover is dominated by Kobresia. Mean annual, January
and July air temperatures at the elevation 3300 m a.s.1.
are -3.5"C, -13.7"C and 6.7"C, respectively. The maximum
annual precipitation is about 1000 to 1100 mm. The
snow cover develops in October. By the beginning of
June, the thickness of snow cover varies from 1 to 5 cm
up to I00 to 120 cm, relative to topography and wind activity.
Depending on geomorphology and microclimatic conditions,
the thickness of permafrost varies from IOto 15 m
to 80 to 90 m on the northern slopes and is absent on the
southern slopes. Temperatures of permafrost vary from
-0.1"C to -0.45"C. There are 10 thermometric boreholes
located on different slopes and aspects within an area of 3
km2. Active layer thickness is determined by temperature
measurements made in boreholes from 3 m up to 70 m in
depth. Several CALM sites are located near a mountain
pass. Initial results of the Circumpolar Active Layer Monitoring
were reported previously (Brown et al. 2000).
Ground temperatures up to 25 m in depth have been
measured since 1973 (Gorbunov and Nemov 1978). Both
interpolation of temperature measurements and excavations
in the moraines revealed active-layer thickness in the
late summer for 1973 and 1974 (Table. I). The table includes
mean active-layer thickness during 1974-1977 for
Cl(K0) and C2(K1) sites. The average increase in mean
annual air temperature for the last I00 years in the central
part of Transili Alatau Range has been 0.02"Clyr. The average
ten-year temperature for 1991-2000 has increased
by 0.3"C in comparison with 1981-90. The greatest increase
of temperature for the same period has been mean
winter (0.4"C), maximum winter (0.9T) and minimum
summer temperature (0.YC). The warmest years for the
last decade of the 20th century were 1997, 1998 and 1999,
when
Table 1. Active-layer thickness near
1973-1977.
fusalykezen Mountain pass for
Excavations Altitude Depth of pits Active layer
and boreholes thickness
m a.s.1. m m
Pits
No 1
No 2
NO 4
No 5
No 10
No 12
Boreholes
C1 (KO)
C2 (KI)
C3a
c7
c11
3337
3334
3334
3337
3336
3334
3337
3328
3337
3334
3335
6.0
4.0
4.3
9.3
9.0
4.2
25.0
14.0
10.3
11.7
13.0
4.5
4.0
3.5
4.0
3.8
3.7
3.2
3.4
4.3
3.5
3.7
mean annual air temperature was higher than the longterm
average temperature by 1 .I"C, 1.49"C and 0.93"C,
respectively. Analysis of meteorological conditions for
summer time has shown that the summer temperatures
and precipitation figures are currently rising (Fig. I).
E 400
300
e
200
100
1930 1940 1950 1960 1970 1980 1990 2000
Figure 1. Air temperatures (a) and precipitation (b) for the summer
time (Central Transili Alatau Range, 2500 m a.s.1.). 1 - annual, 2 -
five-year average, 3 - linear trend.
101

Thus, climatic processes during the 20th century and
especially during the last two decades have a great influence
on the contemporary thermal state of permafrost.
Geothermal observations indicate permafrost warming in
the northern Tien Shan for the last 30 years. The permafrost
temperatures increased during 1973 - 2002 by 0.2 to
0.3"C for undisturbed systems, and up to 0.6"C of those
affected by human activities.
In accordance with interpolation of borehole temperature
data, active-layer thickness showed a significant increase
during the last 30 years from values of 3.2 to 3.4 m
in the 1970s to a maximum of 5.2 m.
The average active-layer thickness for all measured
sites increased by 23% in comparison with the early Seventies.
But compared with the thawed depths for the new
CALM sites (KO, K1, C7) the increase is much greater at
42%. For example, at the CALM K1 site during 1974-
1977, the average active-layer thickness was 3.4 m and
maximum value of 3.5 m. During 1990-2000 at the same
site the average active-layer thickness was 4.85 m with a
maximum value of 5.2 m. At the CALM site KO, the average
and maximum active-layer thickness were 3.2 m and
3.3 m for 1974-1977 and 4.8 m and 5.1 m during 1990-
2000.
As mentioned earlier, 1998 was the warmest year on
record in Transili Alatau since 1964. As a result, in the
following year the ground temperature at a depth of 4.6 m
had increased significantly. For example, the average
temperatures at 4.6 m were -0.2,2.24,0.10,0.06 and 0.04
in 1998, 1999, 2000, 2001, and 2002, respectively. This
deep penetration of seasonal thawing into the ground initiated
the appearance of a residual thaw layer deeper than
5 m at the CALM K1 site. Normally seasonal freezing
penetrates from 4.5 to 5.2 m.
Using data obtained from boreholes, geologic and
geomorphologic data, and knowledge about the extent of
snow cover and vegetation, an attempt was made to calculate
spatial distribution of active-layer thickness for
Zhusalykezen Mountain pass area. An area of 15x2 km*
near the mountain pass was selected for the computations
of depth of seasonal thaw. The deterministic model
(Marchenko 2001) with regular grid spacing of 50x50 m
was used for computation. Figure 2 shows the calculated
spatial changes in active-layer thickness within selected
areas.
Because in high mountain areas permafrost is one of
the dominant factors affecting slope stability, knowledge
about tendencies of active layer development is of great
interest. Model calculations of active-layer thickness allow
for the assessment of surface processes, landscape dynamics
and natural hazards.
REFERENCES
Brown J., Hinkel K. M. and Nelson F. E. 2000. The Circumpolar Active
Layer Monitoring (CALM) Program: research design and initial results.
Polar Geography 24 (3):165-258.
Gohunov A P. and Nemv A E. 1978. The research of temperature of
loose deposits of the high mountain Tien Shan. Cryogenic Phenomena
of High Mountains. Nauka, Novosibirsk: 92-99.
Marchenko, S. S. 2001. A model of permafrost formation and occurrences
in the intracontinental mountains. Norsk Geografisk Tidsskrift
- Norwegian Journal of Geography 55: 230-234.

Active layer temporal and spatial variability at European Russian
Circumpolar Active Layer Monitoring (CALM) sites
G. Mazhitova, "G. Ananjeva-Malkova, **O. Chesfnykh and **D. Zamolodchikov
Institute of Biology, Komi Science Centre, Russian Academy of Science, Syfyvkar, Russia
*Earth Cryosphere Institute, Siberian Division, Russian Academy of Sciences, Moscow, Russia
""Centre for Ecology and Productivity of Forests, Russian Academy of Sciences, Moscow, Russia
Three sites were established in European Russia in the
framework of the "Circumpolar Active Layer Monitoring"
NSF-funded project (Brown et al. 2000). The Bolvansky
site [R24] (68"18'N, 54"30'E) located at the western limit
of the continuous permafrost zone in Europe has four
years of records; Ayach-Yakha [R2] (67"35'N, 64"Il'E)
with seven years of record and Talnik [R23] (67"20'N,
63'44'E) with five years represent "cold" and "warm" extremes
of the range of permafrost conditions in the discontinuous
permafrost zone. This report was developed
as a result of the CALM synthesis workshop held in November
2002 at the University of Delaware.
The sites represent permafrost with temperatures from
-0.5 to -2.0 "C, and sensitive to even decadal-scale climatic
changes (Oberman and Mazhitova 2001). Data from
weather stations show synchronous MAAT fluctuations in
the European Russian Arctic. MAAT is -4.4"C at the
maritime R24, whereas at the continental R2 and R23 it is
-59°C. The difference is due to winter temperatures,
whereas annual sums of positive temperatures (DDT) are
equal (Fig. I).
In spite of similar DDT, differences in thaw depths between
the three sites occurred in all years of record; endof-season
thaw averaged for the period of record was I01
cm at R24 and 102 cm at R23 versus 68 cm at R2. The
differences are due to effects of local factors, first of all,
snow thickness. The two sites with deeper thaw, R24 and
R23, reveal remarkable similarity in thaw values and dynamics
(Fig. 1). Regressions of thaw depths with DDT0.5
using a form of the Stefan solution (Fig 2) clearly indicate
that the same increases in DDT cause much smaller increases
in thaw depths at the cold R2 than at the two
"warmer" sites.
An inter-annual node variability (INV) index has been
proposed by Hinkel and Nelson (in press) to characterize
the consistency of thaw response to thermal forcing. The
site average index is 10% at R23,16% at R24 and 23% at
R2 ranging from 3 to 69%, from 0 to 37%, and from I to
88%, respectively.The values indicate a more consistent
and predictable response of the warmer R23 and R24, as
compared to the cold R2. At R2, the highest INV values
are characteristic for the grid nodes located in a hollow
with very dynamic soil water content and in the area
where bedrock is closer to the soil surface.
0
1300
1200
1100
1000
900
800
700
800
Mean Amual Alr Ternperat urn, "c
1995 1996 1997 lag8 1999 2000 2001 2002
I I I 1 I I I I
Degree Day3 Thaw,%
1995 1996 1897 lssa 1999 2000 2001 2002
Thaw Depth, cm
Figure 1. Climate parameters and active layer depths at European,
Russia's CALM sites.
103

Mazhitova, E. et al.: Active layer temporal and spatial variability at European Russian. ,.
Effects of various landscape factors on thaw depth Oberman, N.G. 1998. Permafrost and cryogenic processes in the
were analyzed. Snow thickness, which is one of the main 540-550, East-European Russian Subarctic. Eurasian Soil Science 31(5):
controls Over permafrost distribution patterns at both re- Oberman, N,G, and Mazhitova, G,G, 2001, Permafrost dynamics in
gional and meso-scale (Oberman 1998) may also work at the north-east of European Russia at the end of the 20th century.
a scale as detailed as that of 100-m CALM grids, if the Norwegian Journal of Geography 55 (4): 241-244.
range of snow thickness within a site is large enough.
Thus, at R24 a talik occurs in a depression under an annually
developing snow drift. Surface topographic forms
(ridges, basins, etc.) from several meters to several dozen
meters in size and a depth to bedrock (R2) are the most
powerful controls on thaw depth. The second important
factor is organic layer thickness, which can be correlated
to thaw depth both negatively and positively, depending
on the topographic form. The positive correlation is observed
at depressions and hollows where soils are saturated
with water. A statistical analysis conducted for R24
showed that a peat soil layer up to 22 cm thick produces a
stronger effect on thaw depths than do vegetation and a
sodden soil layer. Still, thaw depths under major vegetation
classes differ clearly at all three sites. At R2, frost boil
dynamics are correlated to average thaw depths: three
cold summers in sequence with shallow thaw caused new
boils to develop, whereas during the following three warm
summers with deeper thaw all boils became covered with
at least primitive vegetation.
50
30 70' "
f'""";""---C-*
y =0,74)
914
I "I-
1 .I
27,O 29,O 31,O 33,O 35,O 37,O
Sq. root of DA, ckgrees
* R F123 A E4
""
Linear tend (R) .- - -(F123) ..-.-( Fe4)
Figure 2. End-of-season thaw depth versus square
gree days.
root of thaw de-
Surface subsidence was assessed at R2 by repeated
leveling of each grid node. Over a four-year period site
subsidence averaged 3 cm in total. Thus, permafrost vertical
retreat is underestimated if one bases conclusions on
thaw depths alone.
REFERENCES
Brown, J., Hinkel, K.M. and Nelson, F.E. 2000. The Circumpolar Active
Layer Monitoring (CALM) Program: Research Designs and
Initial Results. Polar Geography 24(3): 165-258.
Hinkel, K.M. and Nelson, F.E. in press. Spatial and Temporal Patterns
of Active Layer Thickness at Circumpolar Active Layer Monitoring
(CALM) sites in Northern Alaska, 1995-2000.

8th International Conference on Permafrost Extended Abstracts on Currant Research and Newly Available ~ n ~ ~ ~
Comparative analysis of active layer monitoring at the CALM sites in West
Siberia
E.S. Melnikov, M.O. Leibman, N. G. Moskalenko and A.A. Vasiliev
Earth Cryosphere Institute, Russian Academy
of Sciences, Siberian Branch, Tyumen, Russia
The main objective of the study is to report on recent analyses
of spatial and temporal active layer variations from the
CALM grids within several bio-climatic and permafrost zones
of West Siberia (Melnikov et at. 2001). These sites were established
in the framework of the “Circumpolar Active Layer
Monitoring” NSF-funded project (Brown et al. 2000). In addition
to the short-term CALM data, other terrain information and
longer records are utilized.
The zones covered by the monitoring data are typical tun
dra (sites “Mar&”” and “Vaskiny Dachi”), and northern
taiga (site “Nadym”). Tundra is characterized by variable
landscapes of hilly plains. The northem taiga has a lesscomplicated
landscape structure and generally more subdued
or flatter topography compared with the typical tundra. Distinguishing
features of the three CALM monitoring grids are:
Mane-Sale [R3] (69043’N, 66045’E; 1000-m grid). Marine-influenced,
vast areas of blowout sands, highly dissected
surface, sparsely distributed peat-moss cover.
Vaskiny Dachi [R5] (7001 7’N, 680 54’E; 100-m grid).
Landslide affected, some blowout sands, saline clay near the
surface, highly dissected surface, sparsely distributed peatmoss
cover.
Nadym [RI] (65020’N, 72055’E; 100-m grid). TechnG
genic impact outside the grid, permafrost occurs mainly with
peatlands and frost-heave mounds, rather flat surface, thick
peat and widespread moss. Zonal changes in the thickness d
the active layer are controlled by a combination of factors.
The summer temperature as well as the thickness of the organic
mat and coverage decreases northward. The maximum
active layer in peat increases respectively from 80 cm
in the taiga to 50 cm in tundra, and in sand from 3 m in the
taiga to 2 m in tundra. The difference in thaw rates in various
climatic environments has common features with a slight rise
of thaw rate northward in mineral soils associated with the re
duction of thaw depth.
Active layer landscape interrelation is controlled to a high
degree by the composition and wetness of soils. Maximum
active layer depth is characteristic of sands, especially at the
blowout hollows. In addition to blowout sands, maximum active
layer is found on the wettest landscapes, (1) in northem
taiga, in mires and tundra with peaty hummocks, (2) in typical
tundra, in ravines. Comparison of active layer in different
zones is possible within similar landscape types.
Marre-Sale has beenthesite of long-termobservations
since 1978 (Vasiliev et al. 1998) at several grids and tmnsects.
Minimum thaw depth was observed in 1999 (87 cm)
followingthe low summerairtemperature. Maximum thaw
depth in 1984, and 1995 (133 cm) is in agreement with high
summer air temperature (6.8 to 7.0oC), while peak thaw in
I990 (132 cm) was not connected with summer temperature,
which was moderate (5.8oC), but was probably due to rather
low summer precipitation.
The Vaskiny Dachi site has been active since 1990
(Leibman 2001). The minimum active layer (81 cm) was ob
served in 1992 (transect) and 1997 (CALM grid), and the
maximum was in 1995 (98 and 94 cm, respectively). The
interannual active layer variability does not directly depend
on the air temperature fluctuations, but also on atmospheric
precipitation and its regime, at least for the minimum active
layer depth.
At the Nadym transect (Moskalenko at al. 2001) starting
in 1972, minimum thaw was observed in 1972 and 1992 (I00
cm), while maximum thaw occurred in 1989, 1995 and 1998
(190cm). At the CALM grid since 1997 the maximum thaw
(143 cm) was observed in a year 2002, the year with rather
a moderate air temperature but high summer precipitation.
Minimum active layer occurred in 1999 (125 cm). There is a
correlation (average coefficient -0.6) between the peat thick-
ness and active layer depth at the Nadym CALM grid.
Thaw depths for the tundra CALM gridsatMarre-Sale
and Vaskiny Dachi are compared (Fig. 1). The trend is similar
at sandy sites, though a certain difference in the active
layer depths is observed. Since the climatic situation for both
sites is rather similar and data is taken from the same source
the explanation may be found in surface conditions reflected
in ground temperature. Comparing observed ground temperature
at both sites we conclude that blowing and removal
of the snow in winter from the hilltops of Vaskiny Dachi de
termines a rather low ground temperature, about -7 “C, while
at Marre-Sale ground temperature averages -5 “C.
105

Melnikov, E.S. et 81.: Comparative analysis of active layer monitoring ...
160
60
2 4 6 a
MSAT, OC
I
-
Marre-Sale
blowout
depressions
- IVaskiny
Dachi
blowout
depressions
- - 1
Vaskiny
Dachl, all
sands
h, cm
200
150
100
50
0
Tundra
Northern taiga
' I "1
2 4 6 8 10 12 1
MSAT, "C
Figure 1. Correlation between the mean summer air temperature
(MSAT) and the active layer depth at the sandy surfaces of Marre-
Sale and Vaskiny Dachi CALM grids.
The long-term data from transects and CALM grids for
tundra and taiga are presented in Figure 2. Landscape types
are characterized by specific response to the atmospheric
warming, independent of the bio-climatic zone. A comparison
was made of mires: wetlands with water table near the surface,
and peatlands with thick peat. As follows from the trend
slope on Figure 2, the response of mires (the upper set d
points on the diagram, Fig. 2) to possible warming is more
strongly expressed as compared with peatlands (the lower
set of points on the diagram, Fig. 2).
Thus, temporal and spatial variations of maximum thaw
depthin two bio-climatic zones of West Siberia are under
study. The following conclusions are drawn.
1 Various bio-climatic and permafrost zones in West Siberia
show little or no warming trends during the study period.
In tundra, no trend is noted. In taiga, the average trend is
0.03"C per year.
2 A slight trend can be discerned for thaw depth. In the
typical tundra and mires of the northern taiga of West Siberia,
active layer thickness increases by 0.2 % and
0.5% per year, respectively. During the last several
years, the north of West Siberia was characterized by
drier summers; therefore the possible reason for increased
depth of the active layer is atmospheric precipitation rather
than air temperature changes.
3 Response of the active layer to climate fluctuations differs
for various landscapes. In both bio-climatic zones, peatlands
are most passive due to the existence of a thick insulating
organic layer. In tundra, the most active response
is determined for wet terrains. In taiga, mires are the most
responsive.
Figure 2. Model of the climate warming: correlation between mean
summer air temperature (MSAT) and active layer depth (h). Larger
symbols refer to the CALM grids, smaller ones to transects; black
markers refer to Marre-Sale (the tundra zone), and shaded markers
are for Nadym (the taiga zone).
ACKNOWLEDGEMENT
This report was developed as a result of the CALM synthesis
workshop held in November 2002 at the University d
Delaware.
REFERENCES
Brown, J., Hinkel, K.M. and Nelson, F.E. 2000.The Circumpolar AG
tive Layer Monitoring (CALM) Program: Research Designs and
Initial Results. Polar Geography 24 (3): 165-258.
Leibman, M.O. Active layer dynamics and measurement technique
at various landscapes of Central Yamal. Earth Cryosphere V(3):
17-24 (in Russian).
Melnikov, E.S., Vasiliev, AA., Malkova G.V. and Moskalenko N.G.
2001. Monitoring of the active layer in the Western part of the
Russian Arctic. First European Permafrost Conference, Rome,
26-28th March 2001. Abstracts.
Moskalenko, N.G., Korostelev, Yu. V. and Chervova, E.I. 2001.
Monitoring of Active Layer in Northern Taiga of West Siberia.
Earth Cryosphere V(1): 71-79 (in Russian).
Vasiljev A.A., Korostelev Yu. V., Moskalenko N.G. and Dubrovin VA
1998. Active layer monitoring in West Siberia under the CALM
program (database). Earth Cryosphere
sian).
ll(3): 87-90 (in Rus-
106

8th International Conference on permafrost Extended Abstracts an Current Research and Newly Available 1~~~~~~~~~~
Recent evolution of permafrost in both Becs-de-Bosson and Lona
glacierlrock glacier complexes (western Swiss Alps)
S. Metrailler, R. Delaloye and *R. Lugon
Dept. Geosciences, Geography, University of Fribourg, Switzerland
"University Institute Kurt Bosch, Sion, Switzerland
INTRODUCTION
Geoelectrical and BTS (Bottom Temperature of the winter
snow cover) measurements were repeated a decade after the
first investigations on two neighbouring sites of the Valais
Alps (Switzerland, 4609'N, 7031'E) where glaciers largely
overrode rock glaciers during the Little Ice Age (LIA). The
main objective of the present study is to provide an overview
of the recent evolution of permafrost in these glacierlrock glaciercomplexes.
Two different time scalesassociatedwith
specific processes are involved: (a) the LIA and the impact
on permafrost bodies produced by the glacier advance, and
(b) the last decade of the 20th century and the response d
permafrost to climate warming.
Tenthorey (1993) and Eerber (1994) investigated both
glacierlrock glacier complexes of Becs-de-Bosson (2630 -
2900 m a.s.1) and Lona (2640 - 3000 m ad.) during the period
1989 to 1991, By means of geophysics, these authors
observed that the permafrost distribution within the rock glaciers
had been affected by the glacier advance that had occurred
during the Little Ice Age (LIA).
Most of the numerous vertical DC (direct current) resistivity
soundings which were carried out around I990 were re
peated in 2002, as far as possible at the same places. The
data were supplemented by further geoelectrical and BTS
measurements.
METHODS
Twenty-eight DC resistivity soundings were performed in
2002,of which ten were new ones. Contrary to the I90
soundings, when only the symmetrical Schlumberger array
was used, the dissymmetrical Hummel configuration was
systematically applied for the more recent acquisitions. Such
a configuration allowed lateral changes in the ground resistivity
to be detected. It was difficult to relocate the exact location
of the soundings that were camed out a decade earlier at the
Becs-de-Bosson rock glacier. At Lona, Gerber (1994)
marked the position of the soundings with wooden stakes.
was therefore possible to reiterate exactly the same soundings.
It
A resistivity mapping was carried out in the rooting zone
of the Becs-de-Bosson rock glacier. The apparent resistivity
was investigated at a pseudodepth of about 10-15 m (20 m
inter-electrode spacing, Wenner configuration, 177 measurements).
A BTS map based on 281 measurement points was
drawn for the Becs-de-Bosson rock glacier. A similar task is
planned at Lona.
Finally, 30 minidata loggers (UTL-1) were dispatched an
both sites in September 2002 in order to control the thermal
state of the ground surface during at least one year.
IMPACTS OF THE LITTLE ICE AGE
Becsde-Bosson
The small Becs-de-Bosson glacier had completely disappeared
by the middle of the 20th century. During the LIA, the
glacier covered the upper half of the rock glacier. Several
push moraines were formed.
The BTS map largely fits with geoelectrical data and
therefore permits the discontinuous distribution of permafrost in
the rooting zone of the rock glacier to be illustrated (Fig.1).
The zone where the BTS values are above -2°C points b
the current probable absence of permafrost. Most of this area
was covered by the glacier during the LIA and probably at
least until the end of the 19th century. Colder BTS values indicate
permafrost occurrence. They correspond either to push
moraines or to the intact lower part of the rock glacier. The
foot of the northem steep slope of the Becs-de-Bosson peak
is cold as well.
h a
During the LIA, the Sasseneire glacier almost completely
overrode the Lona rock glacier. In some places, the glacier
appears to have gone beyond the rock glacier extent. Only a
few ice patches remained at the Foot of the Sasseneire headwall
in 2002.
Geoelectric data showed that only pieces of the rock glacier
have continued to exist since the LIA glacier advance.
Today, there is almost no ground ice whatever left in the en-
107

tire glacierlrock complex. However, permafrost was de
tected, notably, in the terminal part of the rock glacier. This
frozen body seems to have been pushed down by the glacier
towards a more sunny area and far away from any source d
debris.
Thiscouldbeinterpreted as a dramaticincreaseinground
temperature and probably a melting of permafrost between
about 4 and 10 m depth.
CONCLUSIONS
The permafrost and ground ice distribution within the two glacierlrock
glacier complexes reflects the glacial history over
the last centuries. The investigated rock glaciers have been
considerably remodelled by the LIA advance of small glaciers.
Some perennially frozen materials were displaced,
while others had melted completely.
Over the last decade of the 20th century, no tangible
evolution of the permafrost geometry, that could be attributab
to changing climatic conditions, was detected by means d
DC resistivity soundings in both glacierlrock glacier complexes.
The permafrost modifications observed at the front of the
Lona glacierhock glacier complex are attributed rather to the
current unfavourable location of this permafrost consecutive 03
LIA glacierlrock glacier dynamic. This melting of the permafrost
had begun a certain time before 1990, as Gerber (1993)
already observed a settlement of about I m in the same area
by comparing aerial photographs taken in 1972 and 1992.
ACKNOWLEDGEMENTS
Figure 1. BTS map on the upper half Becs-de-Bosson rock glacier.
This project was supported by the Valaisian Society of Natural
Sciences, the University Institute Kurt Bosch and the De
partment of Geosciences at the University of Fribourg.
CHANGES BETWEEN 1990 AND 2002
The comparison between the I990 and the 2002 data shows
an evident change for only two of the resistivity soundings.
The other sixteen soundings for 2002 were not significantly
different than those from a decade earlier.
REFERENCES
Gerber E. 1994. Geomorphologie und Geomorphodynamik der
Region Lona-Sasseneire (Wallis, Schweizer Alpen) unter
besonderer Berucksichtigung von Lockersedimenten mit
Permafrost. PhD thesis 1060. Fac. Sciences, Univ. Fribourg,
Switzerland (in German).
Tenthorey G. 1993. Paysage geomorphologique du Haut Val de
Rechy (Valais, Suisse) et hydrologie libe aux glaciers rocheux.
PhD thesis 1044. Fac. Sciences, Univ. Fribourg, Switzerland (in
French).
Figure 2. Vertical resistivity sounding Lo-906 repeated at a decade
interval on the terminal part of the Lona glacierlrock glacier complex.
The two different soundings were performed at the front d
the Lona rock glacier. Similarly, they indicate a decrease d
the resistivity in the uppermost former frozen layer (Fig. 2).
108

Application
lnsfitufe of Physical, Chemical and Biological Problems of Soil Science of the Russian Academy of Sciences, Pushchino,
Moscow region, Russia
INTRODUCTION its application incorrect lot of for land a MV's surfaces. are
available now with the correct regional and simultaneous de
When Digital Elevation Models (DEM) are used for land sur- scription of planar and profile characteristics of land surfaces
face analysis, quantitative characteristics or morphometric (Shary et al. 2002).
variables (MVs) must be used. Physical phenomena, which
could be described by MVs, might be subdivided into the Geomefrica,,andfoms
following 4 groups:
Morphometric conditions of surface run& and soil through- Geometrical - landforms are land surface forms which do not
flow. need to take gravitational
Geometrical landforms. tion. For example, ridge and valley forms are referred to as
Morphometric conditions of thermal regimes on slopes. geometrical forms if they are defined independently from their
Morphometric conditions of altitude zonality, i.e.,the change orientation in gravitational fields. Locally, maximum and
of atmospheric properties with elevation. minimum curvatures describe such forms.
All 4 groups occur in the cryolithic zone.
MORPHOMETRIC VARIABLES
The land surface can be described by regional and local
MVs. In the first case, the influence of remote sites is taken
into account, in the second case it is not. In this study, a
system of 18 MVs of general geomorphometry was used
(Shary et al. 2002).
Morphometric conditions of runoff
The planar land surface is regionally described by maximum
catchment area (MCA) and maximum dispersal area (MDA).
MCA for a given matrix element (square on a map) shows
the maximum area from which material moving downslope
may be collected. MDA describes the maximum area on
which materials moving downslope may be diffused.
The planar dimension and profile of the land surface are
locally described by horizontal (kh) and vertical (kv) curvatures.
The differentiation of land surfaces with respect to kv
and kh allows us to group them into four local landform types
(Troeh 1964).
Algorithms are frequently used which integrate slope gradient
and aspect (GA) to simultaneously describe both planar
and profile characteristics. The most popular formula is the b
pographical index (Beven and Kirkby 1979). This method,
however, tends to approach infinity with GA-0 and makes
APPLICATION OF MORPHOMETRIC VARIABLES
Here we consider the application areas of an MV system in
geocryology but have not considered the morphometric conditions
of altitude zonality and of the thermal regimes d
slopes. The primary possibility concerns the analysis of tunoff
directions and potential flowspeeds (movement of water,
avalanches, glaciers, lava, etc.).
The MCA map reflects both potential and real hydrological
channel networks without distinguishing them (Fig. 1 and 2);
during calculation, the locations and depths of all depressions
were also computed. This MV is of a critical value and de
termines free water at the surface. The places of output of
surface waters and their areas of their origin are well described.
For the definition of potential runoff rates it is necessary b
take the slope profile along the length of the flow path into
consideration. Areas of relative flow deceleration always coincide
with the concave parts of the profiles, while areas d
relative flow acceleration correspond to the convex parts.
The development of cryogenic processes such as thermokarst
and solifluction, the formation of frost mounds and
alas etc. is connected with theoccurrenceofwater in the
soillground. This allows the prediction of position and potential
intensity based on the correlation of the soil/ground moisture
content (and some other parameters) with the MVs. It is important
to note that in some of the terrains from the 18 studied
109

MVs, only the geometrical form provided a strong correlation
with soil moisture.
Already existing cryogenic landforms can be found as a
result of morphometric analysis of DEMs. In general, landforms
of different genesis can be found. Landforms with gee
morphometric structures typical for cryogenic landforms can
be separated from other landforms but only by the application
of all 18 MVs.
Mitusov, A. V and Mitusava, 0. E.: Application of quantitative land-surface.. .
forecast the potential imgation a of territory and to estimate he
hazards.
(2) predict the places of development of cryogenic pro^
esses (thermokarst, solifluction, formation of frost mounds and
alases, etc.) as well as their potential intensity.
(3) reveal cryogenic landforms on DEM and separate
them from other landforms (geological, etc.).
ACKNOWLEDGEMENTS
The authors are grateful to P.A. Shary for providing Analytical
GIS Eco with data and Dr 0.1. Khudyakov for advice in
the geocryology field.
REFERENCES
Beven, K.J. and Kirkby, M.J. 1979. A physically based, variable contributing
area model of basin hydrology. Hydrological Sciences
Bulletin 24: 43-69.
Shary, P.A., Sharaya, L.S., and Mitusov, A.V. 2002. Fundamental
quantitative methods of land surface analysis. Geoderma 107
(1-2): 1-35.
Troeh, F.R. 1964. Landform parameters correlated to soil drainage.
Soil Science Society of America Proceedings 28: 808-812.
1062 1 ho
338 s
106 s
33 9
101
3 dl
1 0%
U 343
0 109
003&1
onlo9
0 0031
OBO!!
0 00034
00OOlU
0 OOOOI
0
unlmownv
Figure 2. Example of a matrix with MCA calculated from the DEM
given in Figure 1. The map of MCA allows to predict the location of
rivers. It follows from comparison between Figure 1 and Figure 2
that the locations of rivers coincides with areas of the large MCAs.
CONCLUSIONS
The application of an extended system ofquantitativeland
surface analysis methods in cryology - except for the morphometric
conditions of thermal regimes on slopes and altitude
zonality - allows us to:
(1) define a direction and potential speed of flow movements
for various substances (water, glaciers, lava, landslides,
snow avalanches, etc.); on this basis it is possible k~
110

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available ~ ~ f
Modeling permafrost distribution in the northern Alps using global
radiation
0. Mustafa, M. Gude and *M. Hoelzle
Department of Geography, University of Jena, Germany
"Glaciology and Geomorphodynamics Group, Department of Geography University of Zurich, Switzerland
INTRODUCTION
In order to reduce natural hazards in alpine areas high
mountain permafrost became an important objective of research
within the last decade. Many investigations were
made in the central Alps. Here, the spatial distribution of
permafrost is well documented because a high number of
field measurements and computer models give reasonable
results. But the adaptation of these findings to the
northern Alps margin is assumed to cause problems due
to the different climatic conditions controlling the permafrost
regime. A regional analysis should especially account
for the higher diffuse radiation compared to direct radiation.
Therefore, the CLOUDMAP model is developed as a
variant of the PERMAMAP model.
THE TEST SITES
The investigation area in the mountains of Wetterstein is
situated at the northern margin of the Alps at 47"25'N and
IO"59'E. It is crowned by Germany's highest peak, the
Zugspitze (2962 m a.s.1.) and the lowest elevation is at
1580 m a.s.1. A 20 m grid is used for model runs at this
site. For calibration and comparison a secondary test site
in the central Alps south of Zermatt (Valais) was chosen
(47"02'N, 7'45'E). The 25 m DEM of this site reaches
from 1675 to 3424 m a.s.1. Climatic parameters with relevance
to the models are listed in Table I.
Table 1 : Climatic parameters of the test sites.
Wetterstein Zermatt
0°C-lsotherme [m NN] 2101 2363
Temp. gradient [WIOO m] 0.60 0.51
Daily sunshine [h] 5.7 6.0
Cloudiness [I1101 6.9 4.1
The Zermatt test site has, for years, been the site of
intensive permafrost research. A great deal of field data
are collected and various computer models were applied
(e.g. Gruber and Hoelzle 2001). In contrast, there is only
little information about the permafrost situation at the
Wetterstein test site (Ulrich and King 1993).
Permafrost-mapping was accomplished by long-term
measurements of BTS and bedrock temperature and
analysis of observed permafrost and permafrost related
features (Gude and Barsch 2003). For spatial extrapolation
of the measurements, GIS-based modeling was undertaken
with two different models, which will be presented.
PERMAFROST DISTRIBUTION MODELS
The PERMAMAP model (Hoelzle 1994) calculates the
spatial distribution of alpine permafrost by applying a statistical
relation of field measurements of BTS (bottom
temperature of snow) with elevation and direct solar radiation
during summer.
To account for the more humid and, therefore, more
cloudy conditions in the northern Alps the CLOUDMAP
model was developed. CLOUDMAP relates the BTS to
elevation and global radiation (Fig. 1). An approach by
Arck & Escher-Vetter (1997) is the basis for the calculation
of global radiation. In contrast to the PERMAMAP model,
CLOUDMAP needs the input of meteorological data for
cloudiness and duration of sunshine.
RESULTS
Both models were applied to both test sites. Due to the inaccessibility
of large parts of the Wetterstein during winter
it was not possible to get enough BTS-data to generate
the statistical relations the models need. Hence, a regional
calibration by climatic parameters was applied to the statistical
relations of the Zermatt test site in order to transfer
them to Wetterstein.
A comparison of the model results for the Zermatt test
site shows only minimal differences. In contrast, the results
for the Zugspitze test site differ significantly (Fig. 2).
The difference concerns most notably the north-exposed
slopes, where the lower limit of permafrost rises about 250
m of elevation.
111

Mustafa, 0. et al.: Modeling permafrost distribution in the northern Alps
........................
..........................
MAAT
Global Radtatlon
Ij
.........................
Input j
...
I
Output
Figure 1. Structure of CLOUDMAP as an insertion to PERMAMAP; fl-3(X) - empirical functions
Figure 2. Modelled permafrost distribution at the Wetterstein test site (glaciers are not considered).
REFERENCES
Arck, M. and Escher-Vetter, H. 1997. Topoclimatological analysis of
the reduction of the glaciers in the Zugspitz region, Bavaria. Zeitschrift
fur Gletscherkunde und Glazialgeologie lX(1-2): 57 72.
Gruber, S. and Hoelzle, M. 2001. Statistical modelling of mountain
permafrost distribution: Local calibration and incorporation of remotely
sensed data. Permafrost and Periglacial Processes 12(1):
69 - 77.
Gude, M. and Barsch, D. 2003 (in print). Assessment of geomorphic
hazards in connection with permafrost occurrence in the Zugspitze
area (Bavarian Alps). Geomorphology,
Hoelzle, M. 1994. Permafrost und Gletscher im Oberengadin. Grundlagen
und Anwendungsbeispiele fur automatisierte Schatzverfahren.
Mitteilungen der VAW 132. ETH Zurich.
Ulrich, R. and King, L. 1993. Influence of mountain permafrost on
construction in the Zugspitze mountains, Bavarian Alps, Germany.
Sixth International Conference on Permafrost, Beijing, Proceedings
1 : 625 - 630.
112

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available ~ ~ ~
The circumpolar active layer monitoring (CALM) program: Recent
Developments
F. E. Nelson, N. 1. Shiklomanov, *K. M. Hinkel and **J. Brown
Department of Geography, University of Delaware, Newark, DE USA
“Department of Geography, University of Cincinnati, Cincinnati, OH USA
**International Permafrost Association, Woods Hole, MA USA
The Circumpolar Active Layer Monitoring (CALM) program (2000). Under the aegis of the IPA’s Permafrost and Global
was established to observe the long-term response of the aG Change Working Group, meetings and discussions devoted
tive layer and near-surface permafrost to changes in climatic to CALM have been held at most of the annual cryosphere
parameters. Initial foci of the program included “data rescue” conferences in Pushchino, Russia and at several of the Ameactivities
and creation of a data archive, building an alliance d rican Geophysical Union’s Fall Meetings in San Francisco.
active field scientists willing to share data, execution of critical
The current phase of CALM, supported by the U.S. National
field experiments, and development of data-collection prote
Science Foundation (NSF), reached its apogee in November
cols. Most aspects of these tasks were implemented by the
mid-1990s. Details are provided in the volume by Brown et
2002, when 35 scientists from six countries attended an
al. (2000).
NSF-sponsored workshop in Lewes, Delaware. This
Data produced from the CALM network have proven workshop provided an opportunity for CALM scientists to
useful in many contexts, including model validation. Owing 13 learn details about research programs at various sites, to dithe
limited observational record at most sites, it is not yet scuss progress and problems in the network, to implement
possible to arrive at definitive conclusions about long-term unified data-analytic procedures, and to plan future activities.
changes or trends in the temperature and thickness of the ac- The CALM network currently includes 130 active sites
tive layer. The few long-term data sets available from high- with 15 participating countries. An additional 20 secondary
latitude sites in the Northern Hemisphere show very sub- sites are cc-located with these primary sites, resulting in a
stantial interannual and interdecadal fluctuations. The spatial total of 150 sites that report active-layer thickness on an ak
variability of active-layer thickness is large, even within gm
nual basis. Of the total, 100 sites are collecting soil and pergraphicareas
of very limited extent (e.g. 1 km2). Most sites
mafrost temperature data; of these, 60 are from boreholes
in tundra environments show a strong, positive relation be
tween summer temperature and the thickness of the active
layer. This relation is weaker in boreal environments. Increases
in thaw penetration, subsidence, and development of
thermokarst terrain have been observed at some sites, particularly
in subarctic regions.
Strong bonds have been forged between CALM and several
international monitoring and global change programs.
CALM is part of the International Permafrost Association’s
Global Terrestrial Network for Permafrost (GTN-P), which
between 3 m and 884 m deep (Smith et al, this volume,
discuss metadata information on GTN-P boreholes). Most
sites in the CALM network are located in Arctic and Subarctic
lowlands, although 20 boreholes are in mountainous regions
of the Northern Hemisphere above 1300 m elevation. A
new Antarctic component is forming and currently contains 13
sites.
The 2002 CALM workshop focused synthesis activities
on the 70 sites in six countries from which gridded data are
also incorporates a network of boreholes for monitoring per- available. The workshop resulted in preparation of seven exmafrost
temperatures (Romanovsky et al. 2002). GTN-P is tended regional abstracts for this volume (Table I). Several
itself part of the Global Terrestrial Observing System (GTOS) other papers report details from these same CALM sites
and the Global Climate Observing System (GCOS), co- (Romanovsky et ai. 2002, Hinkel and Nelson 2003) Other
sponsored by a consortium of international organizations, CALM related papers appear in the Proceedings and
including the World Meteorological Organization and the Inter- Abstract volumes for Mongolia (Sharkhuu), Kazakhstan
national Council of Scientific Unions, among others. Further (Marchenko), Canada (Tarnocai et ai.), and Antarctic
details are provided in Burgess et al. (2000).
(Guglielmim et al.).
Only a little more than a decade into its existence, CALM Extended discussions about CALM’S future directions,
has produced a large body of scientific literature, much of it strategies, and activities took place at the Delaware
reviewed in a monograph-length paper by Brown et al. workshop, culminating in the CALM Workshop Resolution.
113

Nelson F.E. et 4.: The circumpolar active layer monitoring (CALM) program ...
The Resolution consists of the following main recommendations:
To continue the CALM observations and improve the
thematic representativeness of the network;
To improve linkages with other networks and international
programmes such as WCRP (CIiC), IGBP and the proposed
Circumarctic Environmental Observatories Network
(CEON);
To better understand active layer dynamics through ob
servations of borehole temperatures, subsidence, snow
cover, soil properties (moisture, surface organics, texture),
topography and vegetation;
To ensure continued operation, maintenance and en-
hancement of CALM and borehole temperature databases
(GTN-P) and websites;
To provide input for improvement of permafrostdimate
model development, model result comparison, and verification;
To prepare reports to disseminate results of data analy-
ses beginning with the 8th ICOP and a special journal
publication; and
To develop strategies for remote sensing-based methods
of monitoring geocryological parameters over extensive
areas, consistent with the GHOST observation scheme
Table 1. Extended abstracts resulting from 2002 CALM workshop
(senior author listed for abstract in this volume).
1. Nordic Region: Christiansen et al.
2. Lower Kolyma River: Fyodorov-Davydov et al.
3. West Siberia: Melnikov et al.
4. European Russia: Mazhitova et al.
5. West Siberia: Vasiliev et al.
6. Chukotka: Zamolodchikov et al.
7. Modelling: Oelke and Zhang
Romanovsky, V., Smith, S., Yoshikawa, K. and Brown, J. 2002. Permafrost
temperature records: indicators of climate change. Eos,
Transactions, American Geophysical Union 83(50): 589 and
593-594.
ACKNOWLEDGEMENTS
The CALM program work was supported by the U.S. National
Science Foundation under grants OPP-9732051,
0094769, 0095088, and 0225603. Views expressed in this
abstract are the authors' and do not necessarily reflect those
of NSF.
REFERENCES
Brown, J., Hinkel, K.M. and Nelson, F.E. 2000. The Circumpolar Active
Layer Monitoring (CALM) program: research designs and
initial results. Polar Geography 24(3): 165-258.
Burgess, M., Smith, S.L., Brown, J., Romanovsky, V. and Hinkel, K
2000. Global Terrestrial Network for Permafrost (GTNet-P):
permafrost monitoring contributing to global climate observations.
Geological Survey of Canada, Current Research 2000-
E14.
Hinkel, K.M. and Nelson, F.E. in press. Spatial and temporal patterns
of active layer thickness at circumpolar active layer monitoring
(CALM) sites in northern Alaska, 1995-2000. Journal of
Geophysical Research-Atmospheres.

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
K.A.Novotofshaya-Vlasova, D.A.Gi/ichinsky, *A.A. Abramov and **V.S. Soina
Soil Cryology Laboratory, Institute of Physicochemical and Biological Problems in Soil Science, Pushchino, Moscow Region,
Russia
“Department of Geocryology, Moscow State University, Moscow, Russia
** Department of Soil Biology, Moscow State University, Moscow, Russia
Sterile frozen samples were taken from boreholes (Fig. 1)
near the Laptev Sea coast (East Siberia) and in Beacon
Valley (Antarctica) for iiclassical” microbiological analyses.
The examined Arctic sediments were formed about 10400
thousand years ago (Schirrmeister et al. 2002). The sediments
are ice-rich (up to 80-90%) and have a low mean annual
temperature (-10 to -15T). We also investigated ancient
Antarctic permafrost, several million years old (Sugden et al.
1995) taken from boreholes drilled in the region of the Dry
Valleys, which have mean annual temperatures ranging from
-20 to -22°C with ice contents lower than in Arctic deposits
(40%). The magnitude of unfrozen water and total organic
matter content (TOC) in Arctic loam is 33%. In Antarctic
sands, both these values are close to zero, i.e., at levels impossible
to record with presently available instrumental methods.
Special procedures were used to determine the biological
activity in the thawed samples. The number of viable cells
was counted using the plate method. In the model experiments,
the long-term viability of isolated species of both
spore-forming and non spore-forming bacteria was maintained,
for I to 4 months in a diluted nutrient medium, that is
tryptic soy broth and water suspensions at room and lower
(7°C) temperatures.
The communities of micro-organisms inhabiting the ancient
permafrost could be easily recovered. These communities
are chiefly represented by bacteria concentrations varying
from 102 to 107 cellslgdw (Fig. 1).
Taxonomic identification of isolated aerobic bacteria
showed that such biotopes contain the same diversity of
bacteria as modern tundra cryosols. The taxonomic diversity
revealed the tendency to survive predominantly as non
spore-forming, presumably psychrotrophic bacteria able b
grow on nutrient media at temperatures of 4 to 22°C. Increase
in the microbial (=permafrost) age or decrease in temperature
of Antarctic sediments in comparison with Arctic
ones result in the decrease of proliferating cell numbers. This
suggests that part of the permafrost microbial community may
either undergo deep anabiosis, or be preserved in stable, socalled
viable, but non-cultivable forms. As mentionedin an
earlier study (Vorobyova et al. 1997), the ratio between cultivable
and non-cultivable microbial cells might indicate the
physiological response of microorganisms to environmental
conditions.
The ability of bacteria to survive in frozen sediments may
be explained by specific mechanisms of the cell differentiation
into forms that enable bacteria to sustain conditions unfavorable
for growth and multiplication (cold stress, starvation,
etc.). The colony-forming units (CFU) in long-stored cultures
decreased from 108 down to 102. Nevertheless, as may be
seen by electron microscopy, the majority of the cells are intact
and resume their activity without additional procedures in
conditions favorable for growth. Bacteria Bacillus sp., and
Azotobacfersp., isolated from permafrost, are known to pre
duce typical resting forms - spores and cysts in their life cycles.
However, for long-term storage under conditions of
starvation and subzero temperatures, they produce other
forms (Fig. 2). Such specific cell forms as observed in the
stored cultures may be regarded as one of the reversible
dormant types of both, non spore-forming bacteria and bacteria
which are known to produce specific resting cells that can
beformed in permafrost. Probably, this mechanism of survival
in extreme conditions might also work in an extraterrestrial
environment. The investigated deposits may be considered
approximately similar to Martian permafrost which,
according to the most recent data, contains ground water ice
(Mitrofanov et al. 2002). The existence of thermo-resistant viable
bacterial cells with specific ultrastructure (similar to dormant
forms) in terrestrial permafrost is a reason for advancing
the hypothesis that Martian permafrost might contain Life.
115
ACKNOWLEDGEMENT
This work was supported by the Russian Foundation for Basic
Research, grant 01-05-65043.

REFERENCES
Gilichinsky D. 2002. Permafrost as a microbial habitat. Encyclopedia
of Environmental Microbiology,Willey: 932-956.
Mitrofanov I. et al. 2002. Maps of Subsurface Hydrogen from the
High Energy Neutron Detector, Mars Odyssey. Science 297: 78-
81.
Schirrmeister L. et al. 2002. Paleoenviromental and paleoclimatic
records from permafrost deposits in the Arctic region of Northern
Siberia. Quaternary International 89: 97-118.
Sugden D. et al 1995. Preservation of Miocene glacier ice in East
Antarctica. Nature 376: 412-414.
Vorobyova E. et al. 1997. The deep cold biosphere: facts and hypothesis.
FEMS Microbiology Reviews 20: 277-290.
116

8th International Conference an Permafrost Extended Abstracts on Current Research and Newly Available Information
Comparing thaw depths measured at CALM field sites with estimates from
a medium-resolution hemispheric heat conduction model
C. Oelke and T. Zhang
National Snow and Ice Data Center, Crysospheric and Polar Processes Division, ClR€S, University Of Colorado,
Boulder, Colorado, U.S.A.
INTRODUCTION
The Circumpolar Active Layer Monitoring (CALM) program
(Brown et al. 2000) was initiated in the late 1980s
to observe and understand the response of the active
layer and near-surface permafrost to climate change.
This systematic study focuses on active layer processes
and monitoring thaw depth. It currently incorporates
measurements from more than 100 sites involving 15 investigating
countries. The CALM site data were collected
primarily over 100 m x 100 m or I000 m x 1000 m
test areas. CALM data are available from the CALM
website at htf~://~2.~issa.uc.edu/-kenhinke/CALMI.
Here we compare these field measurements of thaw
depth with results from a numerical heat conduction
model.
Heat Conduction Model
A finite difference model for one-dimensional heat conduction
with phase change (Goodrich 1982) is used.
This model has been shown to provide excellent results
for active layer depth over permafrost and soil temperatures
(Zhang et al. 1996) when driven with well-known
boundary conditions and forcing parameters at specific
locations. A detailed description of the model is given by
Goodrich (1982) and Zhang et al. (1996). Our intent is to
model the soil freezelthaw state for the whole Arctic
drainage area. We run the model one-dimensionally and
assume no lateral heat transfer among the 25 km x 25
km grid cells with depths of 15 m. Soil is divided into
three major layers (0-30 cm, 30-80 cm, and 80-1500 cm)
with distinct thermal properties of frozen and thawed
soil, respectively. Calculations are performed on 54
model nodes ranging from a thickness of 10 cm near the
surface to I m at 15 m depth. Details about model settings
and the forcing data sets (surface air temperature,
snow thickness, soil bulk density, soil moisture content)
are given by Oelke et al. (2003). The model is spun up
for 50 years in order to obtain more realistic start conditions
for temperatures of all model layers. The model is
applied to the whole Arctic area drainage comprising
39,926 north polar Equal-Area Scalable Earth (EASE)
grid cells with an area of 24.7~106 km2. We use a daily
time step for the 22 years of our simulation (1980
through 2001). Oelke et al. (2003) show fields of thaw
depth for this area at three-month intervals, together with
the active layer depth and the length of freezing and
thawing periods.
COMPARISON WITH CALM THAW DEPTH
MEASUREMENTS
Arctic Drainage Area
We compare our model results for the year 2000 with
measurements of thaw depth (DT) for the same year,
made at 66 locations in regions of continuous or discontinuous
permafrost. Measurements from Canada,
Greenland, Russia, Svalbard, and Alaska are included.
Figure I shows the comparison for the day when the
measurement was taken in the field, which is usually
close to the annual late summer maximum. Also shown
are bars of +_I standard deviation for 48 grid sites. Modeled
thaw depth is a distance-weighed average between
the four grid cells around the CALM site location.
The comparison between modeled thaw depths
based on 625 km2 averages of temperature, snow
depth, and soil bulk density and a measured thaw depth
on a much smaller scale (1 ha or I km2) should be
viewed as an indication of model performance rather
than a true validation. The modeled values nevertheless
agree reasonably well with the measured values within
their standard deviation, and capture the general picture
of high-Arctic active layers. The high standard deviation
of the grid measurements points to large spatial variability
of thaw depth, even on these small scales.
Grid cells with modeled thaw depths much lower than
measured depths are located in Greenland, Svalbard, or
Baffin Island fjords at lower elevations than those of the
higher, and consequently colder, corresponding model
grid cells. The resulting modeled thawing seasons are
shorter and the thaw depths are shallower. The RMS er-
117

or without seven such sites (marked by diamonds in
Figure 1) is reduced from 32.1 cm to 23.7cm. The comparison
to 70 CALM sites for 1999 shows a similar behavior
as for 2000, with a RMS error of 29.7 cm. lt is reduced
to 26.7 cm without six sites that experience a
strong cold bias adjacent to mountain ranges.
Qelke, C. et ai.: Comparing thaw depths measured at CALM filed sites ...
the analysis in 1995 with rul=0.961 and ru2=0.936. Differences
in U1 and U2 thaw depths are up to 2 cm, due
to the different location and size of the sites, and different
probing dates.
Measured and modeled thaw depths agree quite
well, taking into account the considerable differences in
scales and the large spatial variability measured on the
very much smaller CALM grids.
madeled D,
0.0 0.5 1.0 1.5 2.0
Modeled DT (m)
Figure 1. Modeled vs. measured thaw depth (DT) for the day when
the measured value was observed in 2000. Bars give the *I standard
deviation for CALM sites measured on grids. Diamonds mark
sites adjacent to mountains or high plateaus; therefore, modeled
thaw depths are too shallow.
Local Time Series
The comparison between modeled and measured
thaw depth for the two CALM sites near Barrow (north
slope of Alaska at 71' 19'N and 156' 36'W) shows DT
close to the diagonal at about 0.3 m (triangles in Figure
I). A 22-year time series (1980-2001) of active layer
depth (ALD) for the model grid cells near Barrow is
shown in Figure 2. Only a very small positive linear trend
of 0.029 cmla is calculated. The solid bold line gives
thaw depth (DT) for the day of year the CALM measurements
were taken, and is therefore lower than the ALD.
This modeled thaw depth is compared to measurements
at CALM sites U1 (Barrow, I000 m grid, dotted line) and
U2 (Barrow, CRREL 10 m plots, dashed line). Both
modeled and measured values show an increase in
thaw depth from 1991 to 1994. The model value then
decreases before rising to a maximum in 1998. From
1998 to 2001 there is a steady decline in thaw depth at
Barrow. A correlation coefficient rul of 0.421 is calculated
between modeled and measured thaw depth at U1
(1992-2001). For the site U2, Ru2 is slightly higher at
0.609 (1991-2001). Modeled and measured curves are
almost parallel after 1994, with differences up to 5 cm.
The correlations become much stronger when starting
0.0
1980 1985 1990 1995 2000
Yea
Figure 2. Modeled active layer depth (ALD) at Barrow, Alaska, for
1980-2001, including the linear trend (dash-dotted line). Bold dotted
lines are measured thaw depths (DT) at CALM sites U1
(1992-2001) and U2 (1991-2001), and the solid bold line is the
modeled thaw depth for the actual day of measurement.
ACKNOWLEDGMENTS
This report was developed as a result of the CALM synthesis
workshop held in November 2002 at the University
of Delaware. This study was supported by NASA
through grant NAG56820 and by NSF through grant
OPP-9907541.
REFERENCES
Brown, J., Hinkel, K.M. and Nelson, F.E. 2000. The Circumpolar
Active Layer Monitoring (CALM) Program: Research Designs
and Initial Results, Polar Geography, 24(3): 163-258.
Goodrich, L.E. 1982. Efficient numerical technique for onedimensional
thermal problems with phase change, International
Journal of Heat and Mass Transfer, 21: 615421,
Oelke, C., Zhang, T. Serreze, M. and Armstrong. R. 2003. Regional-scale
modeling of soil freezelthaw over the Arctic drainage
basin, Journal of Geophysical Research 108(D10): 4314,
doi: 10.102912002JD002722..
Zhang, T., Osterkamp, T.E. and Stamnes, K. 1996. Influence of the
depth hoar layer of the seasonal snow cover on the ground
thermal regime, Water Resources Research, 32(7):
2075-2086.
118

8th International Conference on Permafrost Extended Abstracts an Current Research and Newly Available I ~ ~
Coastal dynamics of the Pechora Sea under human impact
SA. Ogorodov
Faculty of Geography, Moscow State University, Moscow, Russia
The so-called Pechora Sea is located in the south-eastern
part of the Barents Sea. The Pechora sector of the
Barents Sea is characterised by loose permafrost coasts.
These coasts are relatively stable under natural conditions
but are rapidly destroyed under human impact. A case in
point is the Varandei industrial area where expeditious
measures to protect industrial buildings and living
accomodation are necessary. Human impact upon the
Varandei area activates coastal erosion because of
improper exploitation that does not take peculiarities of
coastal relief and dynamics into account. Coastal erosion
of the Varandei area brings a threat to the settlement, oil
terminal (Fig. I) and airport.
Active exploitation of the Varandei industrial area and
Varandei Island started in the seventies. Varandei Island
is a typical sand barrier beach in the Pechora sea. The
shoreface of this accumulative form is covered with a
dune belt (avandune) of 4 to 7 m high. Laidas or highwater
surge berms occupy the inner part of the barrier
beach behind the dune belt. Varandei Island was subjectted
to very strong technogenic impacts. The main industrial
base was built and Novyi Varandei, a settlement for
3.5 thousand inhabitants. A well-drained dune belt of the
barrier beach, made up of sand beds with low ice content,
was chosen as a place for the settlement, oil terminal and
storehouses because it seemed to be more stable from
the engineering-geological point of view than the
surrounding swampy tundra lowland.
Construction of the settlement and industrial base,
along the edge of the erosional coastal cliff, was
acompanied by repeated removal of sand and sandpebble
sediments for building materials from the
avandune and beach. This is absolutely wrong for the
zones of wave energy divergence, especially in those that
already suffer erosion.
Within the zone of industrial exploitation, the coastal
bluff and the coastal zone experienced considerable
mechanical deformations of landforms because of
transport, mechanical leveling of coastal declivities and
other technogenic disturbances. A systemless use of
transportation and construction techniques, including
caterpillars, caused degradation of soil and plant cover of
the whole dune belt of Varandei Island. Under conditions
of deep seasonal melting the dune belt, formed of fine
sands, is subjected to deflation and thermoerosion. The
extent and rate of these processes are so great that in
places the surface of the island became 1-3 m lower than
before the period of exploitation began. Deflation hollows
became widespread. Numerous deflation-thermoerosional
gullies were formed in the coastal cliffs. As a result the
cliffs become lower, their homogeneity had been
disturbed, the coastal zone loses sediments and, finally,
the rate of coastal retreat increases.
During the 2000 field season we measured the rates of
deflation and thermoerosion with specially equipped
stations. The averaged data of repeated measurements at
more than 50 reference squares has shown that the
thickness of the sand layer blown away by the wind was
IOto 14 cm in technogenically-deformed territories. At the
same time, eolian accumulation took place in the
territories that were not affected by human activity. At the
“erosional” station, we observed the formation of a new
big gully (up to 4 m deep) in the coastal bluff. Up to 400
m3 of sand were removed from the gully itself and from its
catchment area during the two weeks of snow melting in
June. The coast near the Novyi Varandei settlement is a
119

Orgodov, SA.: Coastal dynamics of the Pechora Sea ...
region of wave energy flow divergence and correspon- ACKNOWLEDEMENT
dingly, formation of sediment flows. As a result, coastal
protection in the area close to the Novyi Varandei settle- This work was supported by INTAS grant no. 1489.
ment caused a decrease in sediment supply to the adjacent
areas and, hence, their erosion.
The height of storm surges increased after an earthdam
and a bridge were constructed in the eastern part of
Varandei Island. The latter is an important factor in coastal
dynamics. Previously, during high surges that corresponded
in time with tides, water flowed partly into branches
and channels, thus lowering the surge height and decreasing
its influence upon the coast.
The erosion rate increased considerably under human
impact in the middle-late seventies. In some years it was
up to 7-10 mlyear. It decreased slightly, down to 1.5-2
mlyear, after the coastal protection near the Novyi Varandei
settlement was constructed. However, it remained high
in the adjacent areas. Recent measurements during 1987-
2000 have shown that the rate of coastal retreat in the
region, outside of the settlement, have increased and
reached 3-4 mlyear (fig. 2), that is twice as high as in
similar regions that have no human activity.
I^ 01 * I ; “FnM”, ”,
-1 20 40 60 80 100 120
L, m - dstance +om the base
Figure 2. Dynamics of the coast near Varandei settlement.
The geoecological situation on Varandei Island is
nearly critical. Industrial exploitation of the territory has resulted
in not only the destruction of the natural coastal
system but also caused considerable damage to human
activities. Several housing estates and industrial constructions
have already been destroyed due to coastal retreat.
This damage will increase every year following the cliff retreat
towards the center of Varandei settlement. Oil terminal,
airport and other industrial objects are endangered. It
is therefore necessary to thorough analyze coastal
morpholithodynamic schemes before natural environments
are disturbed.
Active industrial exploitation of the Pechora region demands
a well-thought-out strategy of territory development
and the finding of proper areas for new constructions. This
negative example from the Varandei region shows that a
well-developed ecologically grounded approach to further
exploitation of coastal regions is necessary. After many
years of investigations in the Pechora Sea region, the
Research Laboratory of the Geoecology of the North,
MSU, has worked out a special methodology of
morpholithodynamic research and created a database on
the coastal morpholithodynamics of this area that will be a
basis for solving both fundamental and applied problems
arising in the course of coastal reclamation.
120

Permafrost processes inducing disastrous effects in engineering
constructions. Medvezhje gas-field, Western Siberia
Y. V. Panova
Earth Cryosphere Institute, SB RAS, Moscow, 119991, Russia, yanusia@mail.tascom.ru
The Medvezhje gas-field is located in the northern part of
Western Siberia. There are three main gas pipelines constructed
above ground, they have diameters of 1420 mm
and 1020 mm. The average gas temperature is more than
15 "C. The permafrost in the area is discontinuously distributed
and consists of sedimentary soil of Quaternary
age (mainly at the Kazanskaya and Salexardskaya sites;
Nedra 1989, Express information 1980). A number of earlier
investigations have been conducted in the area under
study (Galiulin et al. 1997).
The Medvezhje gas pipeline system (MGP) is in need
of reconstruction because of the increasing amount of
deformations that occur. MGP was investigated from the
air and in the field to assess its stability.
Besides the engineering issues, a geocryological map
of 1 :lOO'OOO scale was prepared for the area. Locations of
exploration and the associated findings were marked on
the map and listed in a table. For a small selection see
Table 1. Inspection of the MGP revealed 66 damage sites:
18 were a little deformed, 35 middle deformed and 13
badly deformed.
The most widespread type of deformation (Le. bending
of the pipeline) occurs in a horizontal plane. Usually the
pipeline sags and it results in subsequent ground contact.
An indication for the initiation of pipeline deformation is
the appearance of an apparently superfluous length of
pipeline. The observed pipeline deformations are caused
by different processes (Degtyarev 1974):
temperature deformations;
deformations from inner pressure in the pipeline;
. deformations caused by the incorrect loading of
the pipeline or by other engineering constructions
on and around the MPG;
Deformation or loss of support of the ground underneath
the pipeline also determines the character and degree
of deformation. The latter deformations are influ-
enced by the interaction of the pipeline with permafrost
and geotechnical I geocryological conditions along the
MGP route:
the pipeline embankment or the pipeline itself hinder
surface water runoff;
0 the frozen foundation melts under the thermal influence
of the gas pipeline and the skeleton of the
foundation is compressed;
the pipeline is clamped on watersheds (e.g., hills),
but is without support between such fixed points.
This may lead to differential sagging of the pipeline
under its own weight, and, thus, to incorrect
pipeline loading (Fig. 1).
." ballast-
Figure 1. Schematic of an incorrectly ballasted pipeline.
Various combinations of the above listed processes
have been found.
Deformations, caused by the wrong loading of the
pipeline due to the influence of additional engineering
constructions are:
destruction of the MGP embankment and the subsequent
relocation of the embankment ground;
. part of the pipeline is compressed by the superimposed
road embankment that crosses it (Fig. 2).
Road embankment
Figure 2. Schematic image of the vertical deformation
pipeline as it crosses a highway.
of the gas-
After inspection of the MGP and analysis of the field
data we marked out the general exogenous and geologi-
121

cal processes which cause changes in the MPG location
and result in its deformation.
During the operation of the MGP the bog zones in the
area were found to increase under the thermal influence of
the pipeline, as has been confirmed by air-photos over the
years. If the foundation ground is ice-rich and able to subside,
water often initiates thermokarst processes: Particularly
bog water, temporary or constant brooks, subsurface
water, within or above the active layer, influence the pipeline.
In a number of cases drainage was found to occur
lengthwise along the pipeline embankment, for instance,
in the line trench. For elevated sites, the concentration of
water flow in this trench leads to water inundation into the
rock.
At the beginning of gas extraction holes began to form
in the foundations, by thawing, because of missing heatinsulation.
On ice-rich sites, the thawing was accompanied
by the formation of thermokarst slumps. The most typical
types of subsequent deformations are:
non-uniform pipeline slumps accompanied by bends in
the vertical and horizontal planes;
emergence of the pipeline from water, accompanied by
bends in the vertical and horizontal planes.
The pipeline, initially and most frequently, was built on
ice-rich highland, 2-3 m above the ground. Initially it contained
vertical bends, they slowly straightened due to the
ice-rich ground settling because of ice melt.
Beside the thermokarst processes within the slopes of
a river or brook, water erosion also plays a part in the formation
of settlements.
An enhanced deformation of the MPG occurs at permafrost
sites containing ice wedges due to the melting of
the ice bodies. Investigated sites that developed such
Panova, V.V,: Permafrost processes inducing disastrous effects
thermokarst phenomena are confined to polygonal peat
soil with ice wedges or with ice inclusions. In extensive
peat soil areas the pipeline comes out, in general, into the
drained trench which is supported by local peat-material or
by mineral ground with only a thin layer of peat. Many
sites are deprived of an embankment and the open trench
forms into a pond which becomes a thermokarst center
along the ice wedge. Heat transferred by water-flow forms
an initial trench in the ice wedges. They increasingly melt
forming growing trenches filled with water.
The pipeline itself is a powerful heat source, especially
in the summer time. A broken embankment leads to a
change in albedo and thus, to ground warming. As a result
of the deformation and destruction of the melting peat at
the base it is possible for the pipeline to sink to the foundation
ground owing to the thermal influence of the trench
water on the peat soil, even if the ground contains no ice.
This effect lasts until all the peat from the pipeline base
has been pushed away and firm mineral soil has been
reached.
REFERENCES
”Geocryology of USSR Western Siberia” “Nedra” Moscow 1989.
Degtyarev, 1974. “Instruction on account and choice of designs of
holev with thermal protection in a permafrost” Moscow “VSRlgas”
72.
Galiulin f., lsmailov I., Dubinin P. and Samsonova L.N. 1997. Increase
of efficiency of development of gas deposits of Far North
Moscow “Science” 655
Express information, release 3, Moscow. 1980. Gas industry series:
Geology, drilling and development of gas deposits.
Tablel. Example of MGP data taken in file research, summer 1999
Site No. Geomor- Length (L), amplitude (A) and Filling up the Type of construction method and Degree of
phologic type of MGP
ponds with water preservation of the embankment deformation
type of deformation (m)
round the pipeline
site in horizontal plane in vertical
plane
(m)
flood- -11- there is above ground, middle de
46-2 plain of
destroyed by water embankment
brook
504-1 bog landscape
right bend A-25; -11-
left bend A-2
there is above ground, strongly d
destroyed by water embankment
122

8th International conference an Permafrost Extended Abstracts on Current Research and Newly Available Information
Frozen ground data information: advances in the IPA Global
Geocryological Data (GGD) system
M. A. Parsons, T. Zhang, R. G. Barry and *J. Brown
National Snow and Ice Data CenferlWorld Data Center for Glaciology, Boulder, CO, USA
* lnfernafional Permafrost Association, Woods Hole, MA, USA
BACKGROUND
Permafrost regions occupy about 23.9% and seasonally
frozen ground regions underlie on average of approximately
57.1 % of the exposed land surface in the Northern
Hemisphere (Zhang et al. 2003). Data and information on
frozen ground collected over many decades and in the
future are critical for fundamental process understanding,
environmental change detection and impact assessment,
model validation, and engineering application in seasonal
frost and permafrost regions (IPA DlWG 1998). However,
many of these data sets and information remain widely
dispersed and relatively unavailable to the national and
international science and engineering community, and
some are in danger of being lost permanently.
The International Permafrost Association (IPA) has
long recognized the inherent and lasting value of data and
information and has worked to prioritize and assess frozen
ground data requirements and to identify critical data sets
for scientific and engineering purposes. The groundwork
of the IPA strategy for data and information management
was laid during the 1988 IPA Conference in Trondheim,
Norway. Barry (1988) presented a paper on status and
needs of permafrost data and information and organized
an informal workshop on requirements for permafrost data
and information that attracted more than 100 attendees.
The IPA Data and Information Working Group (DIWG)
was established and a formal workshop was held during
the 1993 IPA Conference in Beijing, China. The DlWG has
subsequently organized workshops in Southampton and
Oslo (1994), Potsdam (1995), and Boulder (1996). The
Global Geocryological Data (GGD) project was initiated by
the DlWG during the Boulder workshop in 1996. During
the 1998 IPA Conference in Yellowknife, Clark and Barry
(1998) provided an update on IPA permafrost data and
information management. The first Circumpolar Active
Layer Permafrost System (CAPS) CD-ROM was published
and delivered to the delegates. The IPA Standing
Committee on Data, Information and Communication
(SCDIC) was established, replacing the former IPA DIWG.
FROZEN GROUND DATA CENTER (FGDC)
To continue the IPA strategy for data and information
management and to meet the requirements by cold regions
science, engineering, and modeling community, the
World Data Center (WDC) for Glaciology, Boulder in collaboration
with the International Arctic Research Center
(IARC) has established a new Frozen Ground Data Center
(FGDC) as a key node in the IPAs GGD system. The IPA
SCDIC has organized two workshops in Boulder (2002
and 2003) on the prioritization and assessment of frozen
ground data requirements and identification of data sets.
The FGDC has expanded access to the 1998 CAPS data
and has augmented data holdings. Data and information
for seasonally frozen ground regions from in situ measurements,
satellite remote sensing, and model outputs are
included. Figure 1 is an example of these new types of
data. It shows modeled maximum frozen ground depth for
non-permafrost regions in the Arctic drainage basin (Oelke
et al. 2003).
Figure 1. Maximum frozen ground depth 1998/99
D
250
225
200
175
150
125
100
75
50
25
0
[cm]
Rmax
The FGDC is distributing a new version of CAPS at the
July 2003 International Conference on Permafrost in Zurich.
The ultimate goal of the FGDC is to serve as a central
access point for frozen ground data and information in
123

the international permafrost community and scientific
community at large.
The FGDC has improved access to existing data
through an online search and order system and availability
in the Global Change Master Directory. The FGDC has
also expanded and updated current holdings with maps of
global and regional permafrost, soil temperature, and soil
classification in a variety of grids and data formats especially
geared to aid the permafrost modeling community.
Publication and distribution of CAPS version 2 marks a
milestone for the FGDC and the IPA. CAPS version 2 is a
compendium of GGD data and metadata currently available
in the FGDC and around the world (see Table 1). It
serves as a snapshot of frozen ground and related data
holdings available as of spring 2003 and enables access
to these data for users lacking ready internet access or
high-speed connections.
Table 1. Highlights of CAPS version 2
IPA Program Data
Metadata from the Global Terrestrial Network for Permafrost's
(GTN-P) Borehole Program and the Permafrost and Climate
in Europe (PACE) project.
Annual thaw depth and other summary data from the GTN-P's
Circumpolar Active Layer Monitoring (CALM) Program.
IPA Cryosols Working Group map of Arctic soils
Metadata, photos, and a specialized bibliography from the
Arctic Coastal Dynamics (ACD) project
Review papers and a specialized bibliography form the
Southern Hemisphere Working Group (SHWG)
Maps
The IPA Circum-Arctic Map of Permafrost and Ground Ice in
several new grids and formats.
Regional permafrost maps from Russia, China, Europe, and
Alaska
Northern Circumpolar Soils Map
Regional soil maps from Russia and China
Frozen Ground
Historical soil temperatures from stations across Russia
Soil temperatures from sites in Alaska, China, Mongolia, and
Antarctica.
Global grids of seasonal frozen ground derived from remote
sensing
Borehole and active layer data from historical sites and other
sites outside the GTN-P purview.
Modelled seasonal frozen ground depth for the Arctic
drainage basin.
Bibliographies
A Comprehensive Permafrost Bibliography current through
Decemeber 2002. (A searchable, up-to-date version is
available online at the FGDC web site.)
A bibligraphy of permafrost and related maps
Specialized bibliographies from the SHWG and ACD.
Other Products
Output from several permafrost or active layer models.
Borehole and active layer data from historical sites and other
sites outside the GTN-P purview.
Dozens of regional and site-specific data sets of permafrost,
soil properties, snow cover, and related parameters.
GGD metadata describing over 100 other data sets available
from investigators and institutions around the world.
Identification of additional data and information from all
participating countries, organizations, and individuals are
requested. Suggestions on data acquisition, management
and distribution are always welcome and encouraged. The
IPA Standing Committee on Data, Information and Communication
will continue to coordinate the GGD and CAPS
activities.
ACKNOWLEDGEMENTS
We thank the IPA Standing Committee on Data, Information
and Communication for coordinating the GGD and
CAPS activities. We especially thank the many data contributors
to this important effort. This work was supported
by the International Arctic Research Center, University of
Alaska Fairbanks, under the U.S. NSF cooperative agreement
OPP-0002239. Financial support does not constitute
endorsement of the views expressed in this article.
REFERENCES
Barry, R. G. 1988. Permafrost data and information: status and needs.
In Senneset, K. (ed.), Proceedings of the 5th International Conference
on Permafrost, Trondheim, Norway, 1988. Vol. 1, Tapir Publishes,
Trondheim: 119-122.
Clark, M. J. and Barry, R. G. 1998. Permafrost data and information:
advances since the Fifth International Conference on Permafrost,
in Lewkowict, A. and M. Allard (eds.), Proceedings of the 7th International
Conference on Permafrost, Yellowknife, Canada,
1998: 181-188.
International Permafrost Association, Data and Information Working
Group, comp. 1998. An Introduction to Circumpolar Active-Layer
Permafrost System (CAPS), in Circumpolar Active-Layer Permafrost
System (CAPS), version 1.0. CD-ROM available from National
Snow and Ice Data Center, Boulder, Colorado.
Oelke, C., Zhang, T., Serrere, M. and Armstrong, R. in press. Regional-scale
modeling of soil freezelthaw over the Arctic drainage
basin. Journal of Geophysical Research.
Zhang, T., R., Barry, G., Knowles, K., Ling, F. and Armstrong, R. L.
2003. Distribution of seasonally and perennially frozen ground in
the Northern Hemisphere. Proceedings of the Eighth International
Conference on Permafrost, Zurich, Switzerland, July 21-25, 2003.
We continue to welcome data submissions for inclusion
on the FGDC web site at http://nsidc.org/fgdc/index.html.

8th International Conference on Permafrost Extended Abstracts an Current Research and Newly Available ~ ~ ~
Mapping of rock glaciers with optical satellite imagery
F. Paul, A. Kaab and W. Haeberli
Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
There is a growing concern about permafrost degradation
as a result of ongoing global warming. In particular, rock
fall events and debris flows from now unstable mountain
slopes are expected (Haeberli and Beniston 1998).
Hence, the great effort to map the distribution of mountain
permafrost in the Alps. Rock glaciers are indicative of
permafrost conditions and can easily be recognized on
aerial photography due to their typical shape. Time series
analysis of the derived orthophotos allows the calculation
of flow fields and mass changes (Kaab 2002) but is not
practical for large area mapping on a small scale.
The potential of high-resolution panchromatic satellite
sensors to detect rock glaciers has been analysed in order
to validate the results of permafrost distribution models at
large scales and in remote areas. The investigated sensors
range from I5 m spatial resolution ETM+ to the 1 m
lkonos sensor. The sensors also differ in the spectral
range and ground area covered as well as costs per km2
(Table 1). In general costs increase with higher resolution
while the area covered decreases.
Table 1 Some characteristics of the pan sensors examined.
-
ETM+ SPOT IRS-IC lkonos
Resolution [m] 15 10 5 1
Spectral range [mm] 0.52-0.9 0.51-0.73 0.5-0.75 0.45-0.9
Area covered [km] 180x180 60x60 70x70 11x1 00
Costs per km2 [US$] 0.02 0.33 0.51 ca. 30
In Figure 1 the comparison of the three pan bands from
ETM+, SPOT and IRS-IC is shown together with a ground
based photo of a complex rock glacier system near Piz
d’Err (Crisons). The ETM+ sensor (Fig. la) captures the
rock glaciers quite well as the sensor is also sensitive in
the near infrared where the high reflection of vegetation
contrasts well with the low reflection of bare rock. However,
details of the ridge structure are not visible and the
spatial resolution is thus, not appropriate for a clear identification.
In the SPOT image (Fig. I b) the typical ridges become
visible and identification and large area mapping of rock
glaciers seems feasible. However, the outline of the lower
(relict) part is difficult to recognize as the reflections from
the surrounding meadow and bare rock are quite similar in
this spectral range.
The IRS-IC image (Fig. IC) reveals only few new details,
as compared to SPOT, but seems to be more noisy
(nominal resolution is 6.8 m). The surrounding meadow is
slightly brighter than seen with SPOT although the spectral
range covered is quite similar.
A comparison between SPOT and lkonos imagery
(draped over a DEM with 25 m spatial resolution) is shown
in Figure 2 (view to south) for the comparable small Gigli
rock glacier (near Sustenpass, Uri). Although some of the
rock glacier ridges are visible in the SPOT imagery (see
inset), it is by no means comparable to the level of detail
visible from Ikonos. Here also, the typical larger rocks at
the rock glacier surface and the very fine debris at the
front can be discerned. A DEM allows the creation of 3D
perspective views which enhance the identification of further
rock glaciers (white arrows). A large number of
breaches from former debris flows across historic (1850)
moraines can be identified as well.
The regions of likely and probable permafrost are indicated
as high-lighted areas in Figure 2. The model simu-
lates BTS, a threshold of -2.0(-3.0)
OC is chosen to assign
regions with likely (probable) permafrost (Gruber and
Hoelzle 2001). It seems that the rock glaciers originate in
regions with likely permafrost but extend much more further
down into permafrost free terrain. This has also been
observed for many other rock glaciers (Frauenfelder et al.
2001).
We propose that contrast enhanced SPOT pan imagery
can be used to identify and map rock glaciers for
large and remote areas at comparably low costs. The IO
m spatial resolution seems appropriate for validation of
permafrost distribution maps obtained from 25 m spatial
resolution DEM data. For further detailed studies of interesting
regions the I m resolution lkonos imagery is a
valuable alternative to aerial photography but about three
times more expensive.
125

Paul, F. et. al.: Mapping of rock glaciers with optical satellite imagery
REFERENCES
Frauenfelder R, Haeberli W, Hoelzle M. and Maisch M. 2001. Using motely sensed data. Permafrost and Periglacial Processes 12 (I):
relict rockglaciers in GIS-based modelling to reconstruct Younger 69-77.
Dryas permafrost distribution patterns in the Err-Julier area, Swiss Kaab, A. 2002. Monitoring high-mountain terrain deformation from re-
Alps. Norwegian Journal of Geogra-phy 55 (4): 195-202.
peated air- and spaceborne optical data: examples using digital
Haeberli, W. and Beniston, M. 1998. Climate change and its impacts aerial imagery and ASTER data. ISPRS Journal of Photogramon
glaciers and permafrost in the Alps. Ambio 27 (4): 258-265. metry and Remote Sensing 57 (112): 39-52.
Gruber, S. and Hoelzle, M. 2001. Statistical modelling of mountain
permafrost distribution: local calibration and incorporation of re-
126

Morphogenetic classification of the Arctic coastal seabed
Y. Pavlidis and S. Nikiforov and *V. Rachold
P.P. Shirshov Institute of Oceanology, Moscow, Russia
*Alfred Wegener Institute for Polar and Marine Research, Pofsdam, Germany
INTRODUCTION
A precise and standardized classification on the variety of relief
forms of the Arctic shelf is essential for both scientific and
applied purposes. The evolution of the Arctic coastal zone,
which is controlled by the interactions of modern wave and
ice factors and the influence of numerous glaciations and
large-scale sea level fluctuations, significantly differs from
other coastal areas of the World Ocean. Glaciations have left
coastal traces in the western part of the Russian Arctic,
whereas its eastem part was almost completely drained during
glacial periods and the subaerial paleorelief is characterized
especially by cryogenic forms: thermokarst, theme
abrasion, thermodenudation and ctyodislocation.
Earlier classifications of the coastal seabed relief suggested
by lonin et al. (1990) for the World Ocean or by
Shepard (1977) according to the peculiarity of geological
structures or sedimentation processes could not characterize
the variety of forms and regional features in the Arctic zone.
The classification presented in this paper is based upon scientifically
evidenced information on morphology, origin and
age as well as geological and neotectonic characteristics d
the seabed relief. This kind of simultaneous analysis of complex
of interactive factors can be called a “morphogenetic”
approach.
The investigations are based on the results of long-term
research with the application of modern equipment including
narrow-beam and multi-beam echo sounders, Parasound,
side-looking radars, bathymetric maps and the analysis d
numerous sediment samples. According to the Science and
Implementation Plan of the Arctic Coastal Dynamics (ACD)
project (IASC 2001), the classification will be formulated for
integration into a circum-Arctic coastal Geo Information System
(GIs).
RESULTS AND DISCUSSION
The seabed evolution is determined by both modem and
paleogeographical processes such as vertical tectonic and
glacio-isostatic movements. Among the exogenic and en
dogenic coastal processes it is possible to allocate active
processes, which directly contribute to the formation of the
shelf relief, and passive processes, which predetermine the
display of the active processes and direct the course of their
development. Active processes can be modern (hydrogenic,
gravitational, etc.), paleogeographical (glacial, erosional, etc.),
and in some cases anthropogenic processes. Structural
forms, with few exceptions, are related to passive morphogenetic
factors. By using such an approach it is possible b
designate three major categories of the seabed relief: structural,
structural-sculptural and sculptural. This kind of subdivision
is relative in many respects because the majority of relief
forms are caused by both endogenic and exogenic facto
andprocesses. It is necessary todeterminetheprevailing
value of various factors of concrete seabed forms. A certain
cannot be avoided subjectivity in allocating structural and
sculptural forms of a relief, and in their morphogenetic division
into active and passive factors. The need to characterize the
relief “layer by layer” arises inevitably when specialized
geomorphological maps have to be produced and a database
to be used in a GIS environment within in international network
has to be designed. As an example the following layers
could be used: first layer - endogenic background; second
layer - relief forms linked to the action of paleogeographic
factors; third layer - modern relief forms caused by the action
of endogenic processes.
Typical structural forms are large seabeds predetermined
by geological structures and created by endogenic
processes, Le., folded andlor fault-fracture tectonics,
volcanism, vertical movements of the Earth’s crust, etc.
Geological structures underlie all forms of the shelf relief and
determine the position of large relief forms, such as synclinal
underwater depressions, anticlinal and brachyanticlinal
underwater raisings, monoclinal plains, flexure ledges etc.
This large group can be further divided into two genetic types
and some subtypes. (1) The tectogenic type includes plicated
forms, Le., anticlinal, brachyanticlinal, synclinal, brachysynclinal,
monoclinal, uniclinal; disjunctive, i.e., fault and
grabens, horsts, fracture-blocks; and non-broken structural
dislocations, i.e., subhorizontal and inclined plains. (2) The
volcanogenic type includes magmatic and mud volcanic
subtypes. Usually relief forms created by volcanic
processes like lava domes and mud volcanic cones
127

Pavlidis. Y. et al.: Arctic coastal seabed..
(Norwegian Sea) are located in water depthsofmorethan
200 m.
In our morphogenetic classification, structural-sculptural
forms are assigned to a category of transitive, ofien relict formations.
These forms can be divided into the following sub
types: (1) erosion-fectonic, Le. , structural-erosive with fluvial
or glacial relief, structural-erosive-gravity created by turbidity
currents, (2) glacial-tectonic, ;.e., structural depressions in
internal basins of fjords and covered by glacial-marine deposits,
structural swells with accumulative superstructure, structural
forms with a surface of glacial ablation corresponding ID
fjords and adjacent underwater trough valleys, structural
forms with glacial ablation and accumulative marginal troughs
and anticlinal swells blocked by moraines.
Exogenic subaerial and subaquatic processes are the
most active ones. The sculptural relief of the seabed is created
by destructive and accumulative processes such as
ablative and accumulative activity of Late Quaternary glaciers
on the shelves of the marginal and internal seas of polar
and subpolar climatic zones. Glacial forms are most typical
amongtherelictexogenicsculpturalreliefformswithin the
limits of the glacial shelf in the Arctic. The sculptural relief
forms can be divided into seven types and several subtypes:
This study was supported by INTAS (project number 2001-
2332).
(I) glaciogenic, Le., glacial-ablative, glacial-accumulative and
glacial-dynamic, (2) cryogenic, ;.e., thermokarst relict forms
such as extended underwater depressions and hollows REFERENCES
documented by echosounding and acoustic profiling in the
Laptev, East Siberian and Pechora Seas; the origin of these IASC 2001. Arctic Coastal Dynamics (ACD). Science and Implenegative
relief forms is connected with the thermal destruction mentation Plan, International Arctic Science Committee, Oslo,
April 2001.
of ice lenses and wedges during the late glacial transgreslonin,
A., Pavlidis Yu. and Yurkevich, M. 1990. Relief of arctic eastsion,
the size ranges from 2 to IOm depth, 10 to I00 m width ern shelf of Russia and its classification. In A Aksenov (eds),
and extension of 3 to 4 km depending on the thickness of the Geology and geomorphology of shelf and continental slope.
lenses of repeatedly-cavern-load ice; in the Chukchi Sea, the Moscow, Nauka: 24-50.
relict thermokarst relief forms are large depressions which are Shepard, F.P. 1977. Geological oceanography. N.Y.
covered by a 7 to IOm thick Holocene sediment layer;
cryodislocative relief forms represented by folds and frost
mounds have a local distribution in the Arctic, for example
Western Yamal Peninsula, (3) hydrodynamic, ;.e., wave
abrasion forming modem cliffs and benches and paleo underwater
terraces, wave accumulation with ancient underwater
accumulative forms and coastal bamer islands, spits,
beaches, etc., abrasive-accumulative processes forming
modem underwater coastal slopes, mainly subglacial accumulative
and tidal forms such as sandy waves and ridges,
tidal troughs, step tidal benches, etc., (4) torrentogenic, Le.,
accumulative and erosive as, for example, in the Bering
Strait where, due to strong quasistationary currents, a subhorizontal
plain lacking modern deposits is formed in the central
zone while as a result of deposition, underwater cones
are generated in the adjoining southem Chuchi Sea, (5) fluvial,
ire., fluvio-glacial and potamogenic forms with icemarginal
valleys andchannelsofpaleo-rivers,ancientand
modem deltas, river mouth bars, etc., (6) gravitational, Le.,
rockfall-landslides, turbidity currents, and (7) modem anthropogenic,
Le., constructive forms connected with ground spoil,
artificial islands, basements of port constructions, oil derricks
etc., destructive forms with artificial cuts, port channels, waterways
in gulfs and other near-port constructions within
shallow water.
CONCLUSIONS
The Arctic shelf can be described as a zone witch evolved in
the come of long-term endogenic and exogenic interactions.
The initial structure was created by the basement surface
which was and is constantly reworked by a complex of paleogeographical
and modem processes within various glacial
and interglacial periods and sea level fluctuations. The classification
suggested in the present article is based on scientifically
evidenced data on morphology, origin and age, and on
the geological and neotectonic features of the seabed relief,
and it also takes into account a multitude of natural factors and
their interactions and evolution in time. Three types of major
relief-forming processes were distinguished and further subdivided:
structural, structural-sculptural (transitive), and sculptural
(formed by modem and paleogeographical exogenic
processes).
128
ACKNOWLEDEMENT

Near-surFace ground temperatures and permafrost distribution at
Gornergrat, Matter Valley, Swiss Alps
S. Philippi, T. Hen and 1. King
Institute of Geography, Justus Liebig University, Giessen, Germany
INTRODUCTION
Due to substantial research during the past fifteen years,
permafrostdistributionoftheMatter valley is known on an
overview scale. However, the currently available permafrost
distribution models do not consider local soil properties which
can vary significantly within short distances in high mountain
environments. Besides other local factors such as air
temperature, solar radiation and snow cover, changes in
substrate and moisture conditions strongly influence the
ground thermal properties (Williams and Smith 1989) and
consequently, the distribution pattern of mountain permafrost.
The zone of discontinuous permafrost is characterized by
rather warm ground temperatures insignificantly below zero,
which makes permafrost at its lower limit very susceptible b
increasing air temperatures. Active layer thickening, rising d
the lower limits of permafrost and permafrost reduction by
basal melting could be possible responses to long-term
climate changes (Williams and Smith 1989).
The current investigations will provide quantitative insights
into ground thermal regimes and will correlate measured soil
temperatures with different soil properties and their possible
effects on the discontinuous permafrost pattem.
The Gomergrat area is located between 2700 and 3200
meters a.s.1. at the end of the Matter valley surrounded by
some of the highest peaks of the Alps. The continental
climatic conditions cause a high glacier equilibrium line
associated with a periglacial belt of large vertical extent.
According to King (1996), discontinuous permafrost can be
expected in an altitudinal range between 2800 and 3500
meters a.s.1. and is indicated by the presence of numerous
active rock glaciers.
SELECTED STUDY SITES
The east-west running crest between Gomergrat (3135
meters a.s.1.) and HohGlli (3286 meters a.s.1.) divides the
study area into a northern part called ‘Kelle’ and a southern
part named ‘Tuff. Both areas are located within the zone d
discontinuous permafrost showing a rich periglacial
morphology such as solifluction lobes, patterned ground and
rock glaciers. However, the local permafrost pattem is not
yet known.
Shallow ground temperature measurements are performed
at eight vertical profiles differing in substrate (cf. Table 1) and
moisture conditions. The measurements are carried out in
order to obtain an insight into the various reasons for the
variability of ground thermal properties over short distances
within the zone of discontinuous permafrost.
In terms of local soil characteristics, each of the four
instrumented profiles in the Tuft corresponds to a profile in the
Kelle. The substrates of location T (Tuft) I and K (Kelle) I
show a rather high soil moisture content, T2 and K2 were
installed in substrates with moderate water content, whereas
T3 and K3 are comparatively dry locations. All sites show a
sparse vegetation cover of moss, lichen and grass and are
located at approax 3000 meters ad. (cf. Table I).
According to Harris and Pedersen (1998) the differing
thermal properties of blocky materials seem to explain
permafrost occurrences below the regional lower limit. In
order to comparethethermalregimesofsoilsandcoarse
debris two additional locations (K4 and T4) within the surface
layer of active rock glaciers were chosen.
Meteorological data are available for the ‘Kelle’ and the
permafrost monitoring site at Stockhorn plateau (3410 meters
a.s.1.) close to the research area.
GROUND TEMPERATURE MEASUREMENTS
Measurement setup
In August 2002 eight data loggers with an accuracy of 0.02
“C were installed, measuring shallow ground temperatures
dierent levels (cf. Table 1). The measurement interval was
set at 15 minutes during summer and autumn, while from
October on hourly recordings are made. So far, data are
availablefromtheendofAugustuntiltheend of October
2002.
129

Philippi, S. et al. 1 Near-surface ground temperatures and permafrost ...
Table I. Information on the location of near-surface ground measurements at Gornergrat.
Profile Altitude depth of sensors location soil type humus within
(until 100 cm)
upper horizon
m cm cm %
TI 3005 2, 25, 50, 100, 160 moist plain 0-15 loam 9.7
15-100 sandy loam
T2 3000 2, 25, 50, 100, 180 solifluction lobe 0-100 loam 15.0
T3 3000 2, 25, 50, 100, 150 knoll 0-100 sandy loam 4.5
T4 3010 50,120,190,255 active rock glacier
K1 2950 2, 25, 50, 100, 150 moist plain 0-1 5 loam 9.5
15-100 silt loam
K2 2950 2,20,45,70,120 plain 0- 40 sand 2.3
40-100 loamy sand
K3 3005 2, 25, 50, 100, 140 knoll 0-1 00 sandy loam 6.3
K4 2950 100, 160, 220, 280,360 active rock glacier
P reliminary results
The observed smaller temperature ranges at the moist profiles OUTLOOK
K1 and TI compared to the drier locations, could be
explained by the low thermal conductivity of water. This To verify the local permafrost pattern, measurements of the
effect is shown in Figure 1, comparing the measured ground Basal Temperature of Snow cover (BTS) and further soil
temperatures at the northern site KI (moist) with the south- temperature readings will becarried out before the I COP
exposed site 13 (dry).
2003 and will be discussed in the poster presentation.
The study area will be visited during one of the postconference
excursions in July 2003.
REFERENCES
12.Y 13.Y 14.Y 15.Y l6.l 17.9 18.9 199 20.4 2l.q 22.9 23.9 24.9 25.9 26Y
h i e
Figure 1. Near-surface ground temperatures in 2 cm and 25 cm
depth in the profiles K1 and K2 (12.09.-26.09.2002).
Harris, SA. and Pedersen D.E. 1998. Thermal Regimes Beneath
Coarse Blocky Materials. Permafrost and Periglacial Processes
9: 107-120.
King, L. 1996. Dauerfrostboden im Gebiet Zermatt - Gornergrat -
Stockhorn: Verbreitung und permafrostbezogene Erschliessungsarbeiten.
Z. Geomorph. N.F., Suppl. Bd. 104: 73-93.
Williams, P.J. and Smith, M.W. 1989. The Frozen Earth. Cambridge:
University Press.
Differences in aspect due to solar radiation mainly affect
the surface ground temperatures (cf. Figure 1). W&
increasing depth local soil properties - especially moisture
conditions I seem to have a major impact on the thermal
regime. Apart from the surface ground temperatures, which
are considerably higher at the southern location, the two moist
profiles show similar temperatures and temperature ranges at
all other measured levels - independent of their aspect.
The temperature amplitude at all measured levels is
significantly greater within the rock glaciers. The phase lag
between surface and soil temperature changes increasing
with depth is missing at the rock glacier profiles. Similar the to
soil locations, the coarse blocky layers also show a
decreasing amplitude with depth, but without any time delay.
Rapid air movement through the gaps seem to be the main
process of heat transfer (Harris and Pedersen 1998) and
would explain the direct response of the temperature changes
within the block layers to a change in air temperatures at the
surface.
130

8th International Conferen~e an Permafrost Extended Abstracts on Current Research and Newly Available information
Results of the engineering geocryological mapping of the Pechora Sea
arctic coast (the key site of the Varandey Peninsula)
A. A . Popova
Industrial and Research lnstifute for Engineering of Construction, Moscow, Russia
INTRODUCTION AND STUDIED AREA
Natural conditions are complex in Arctic Sea coastal areas.
At present, it is necessary to study this area in detail in connection
with oil and gas recovery and all the associated
pipework.
The studies were performed in the Pechora Sea coastal
band with a width of 8-10 km: from the Varandey Cape in the
west to the Khaipudyr Bay in the east. The average annual
air temperature is -5.6 'C. The studies were performed on
the laida (absolute elevation 0-3 m) and the first sea terrace
(absolute elevation 515 m). The sea terrace surface is complicated
by numerous shallow lakes and sedge-moss bogs.
Laida is a low part of the coast and subject to flooding during
tides and storms. Saline sea water penetrates, through the
river valleys, up to IOkm deep into the continent.
The thickness of the Quaternary sediments is 40-50 m in
the area under study. They are represented mostly by sandy
and clayey soils in the first terrace and laida, respectively.
The Varandey Peninsula is located in the zones of continuous
and discontinuous permafrost with a total thickness d
about 250 m. The average annual temperature of the perennially
frozen ground (PFG) varies from 0 to 4 'C (Geocryology
of the USSR 1988).
On the first sea terrace (>95% of the area), PFG is continuous.
Part-through taliks (zones of unfrozen ground) with a
thickness of 5-10 m are confined to small lakes and rivers.
Through taliks are located below large lakes and rivers. In the
first sea terrace, PFG has a thickness of 40 to I00 m and is
underlain by cooled salinized ground with lenses of saline
water (cryopegs). Cryopeg salinity reaches 40-50 gll. The
average annual temperature of the ground is the lowest (from
-2 to -4 'C) at hillocky peatbogs and in areas without vegetation.
The temperatures are the highest (from -0.5 to -2 'C)
within wide swampy runoff bands, khasyreyas (thermokarst
depressions), and the valleys of small streams.
On the coastal part of the area (laida) PFG is discontinuous.
The upper section (to a depth of 34 m) is composed d
cooled ground at a positive temperature. Cooled salinized
ground at negative temperatures (from -1.5 to -3 OC) composes
the lower section. The freezing point of silty clay at a
salinity of 2% is -3 OC. Therefore, frozen ground occupies
only 50-90% of the laida area in spite of a continuous distribution
of ground temperature from 0 to 4 'CC. The physical
properties of a cooled ground correspond to those a thawed of
ground. Therefore, we canstatethat PFG is discontinuous
within the laida. The surface of the frozen ground is also bro
ken by part-through and through taliks below saline rivers and
lakes. The thickness of part-through taliks can exceed 50 m.
Physical geological processes and formations, including
cryogenic heaving, thermokarst and thermal erosion and p
lygonal microrelief are widespread within the first sea terrace.
Coastal erosion is observed at the high cliffs of the first sea
terrace.
METHODOLOGY
Engineering geocryological zoning is the main method used
in this study. It is performed according to the genetic principle
and is based on landscape zoning. Topography, vegetation,
and surface drainage are described for each landscape. The
landscape zoning has been performed using data from field
studies and the deciphering of aerial photographs. Geocryological
areas were distinguished as a result of this zoning.
Genesis and lithology of subsurface sediments, temperature,
thickness and the occurrence of frozen, cooled and thawed
ground and cryogenic processes, were determined for each
area.
The performed studies included the drilling of 25 boreholes,
borehole thermometry, geophysical works (VES and
electric profiling) and the analysis of ground composition and
properties to a depth of I5 m.
RESULTS AND DISCUSSION
The map of engineering geocryological zoning of the
Varandey Peninsula coast was constructed based on the
performed studies. Figure I is only a fragment of this map.
The zoning shows the relationship between the geomorphological
level, genesis, lithology, landscape and average
ground temperature.
Three typical ground sections were distinguished de
pending on the salinity: (I) the entire section (to a depth of 15
m) is composed of salinized ground; (11) non-salinized and
salinized ground occur in the upper (to a depth of 34 m) and
lower sections, respectively; (Ill) non-salinized and salinized

sections, respectively. During the zoning it has been estab the upper section. As a result, although all grounds have
lishedthatgeocryologicalconditionsdependdirectlyonland-negativetemperaturestheyexhibitthepropertiesofthawed
scape conditions. Genetic types, lithology, temperature of the soils. Slightly salinized ground also occurs in the bottom secground
and landscape type are shown on the map for each tions of the first sea terrace.
distinguished area, the shading of each area corresponds b
the average annual temperature of the ground.
The section types with a different distribution of salinity ACKNOWLEDGMENTS
(see above and Fig. 1) are shown by independent indices (I,
II, Ill). This work supported was by the Russian
Program and INTAS grant 2332.
CONCLUSION
REFERENCES
The area under study includes continuous and discontinuous
permafrost. Thawed ground and part-through taliks are local. Geocryology of the USSR. European Part. 1988. Moscow: Nedra (in
The presence and non-uniform distribution of salinized cooled Russian).
Rivkin F., Popova A., and Eospa A. 2001. The Permafrost Condiground
are responsible for specific geoc~ological
tions along the Pechora Sea Coast (the Varandey Peninsula,
in this area. This is especially evident within laidas, where
European North of Russia), In Proceedings of Ist European
the frozen ground is discontinuous because of high salinity in Permafrost Conference, Rome: 38.

8th International Canference on Permafrost Extended Abstracts on Current Research and Newly Available Information
An active rock glacier in the southernmost permafrost environment of the
Alps (Argentera Massif, Italy): four years of surface ground temperature
monitoring
A. Ribolini
Dipartimento di Scienze della Terra, University of Pisa, Via S. Maria 53, 56126 Pisa, Italy; ribolini@dst.unipi.it
The Argentera Massif corresponds to the south-ernmost
permafrost environment of the European Alps, being
only 50 km far from the Mediterranean Sea (Fabre and
Ribolini 2003). This alpine sector is highly sensitive to
environmental changes as demonstrated by last century
glaciers disappearing. The surface ground temperature
(SGT) is used in the analysis of the relationship between
the near surface soil temperature and the nonconductive
seasonal soil heat transfer processes, forcing
permafrost growthldegradation (Kane et al. 2001).
Four years-ground thermal regime at the snow-ground
interface of the Portette rock glacier will be analysed by
means of hourly monitoring of soil temperature using a
miniature digital data logger (DTL). In this rock glacier,
located in the south-eastern sector of the Argentera
Massif, previous geoelectrical sounding indicated the
presence of permafrost between 5-20 m depth (Fabre
and Ribolini, 2003).
SGT data cover the November 1998-November 2002
period (Fig. I). Four onsets of autumn freezing and
spring-summer thawing periods are recognizable. During
winter months the temperature remains well below
O”C, except for the 2000-2001 winter, when it did not
descend below -2°C. The October-May SGT reached
the lowest value during 1998, when the highest annual
“cold content” (degree day of frost, DDF) was experienced.
The winter records show both a regular temperature
decreaselincrease (1999-2000, 2000-2001)
and a .temperature trend characterized by several
“spike-like” anomalies (I 998-1999,2001-2002).
To better understand the ground thermal behavior,
the hourly temperature change DT (“Clh) at the DTL
depth (25 cm) was calculated (Fig. 2). Four seasonal
thermal regimes can be identified: summer thawing
(ST), zero curtain (ZC), autumn-winter freezing (FR) and
snow melt (SM).
15 1999-00 2000-01 2001 -a2
10
5
0
-5
-i 0
-1 5
L’
Snow cover
perlod
-
Snow cover
period
Figure 1, Daily air (grey) and Surface Ground Temperature
The large hourly temperature change characterizing “Zero Curtain Effect” (ZC), i.e. the compensation of the
the ST regime can be considered to represent the rapid upward heat loss associated with the drop in air temperaresponse
of the active layer to weather fluctuations. tures by the freezing latent heat of the soil. The hourly
The ground freezing thermal behavior is characterized temperature change of the autumn-winter FR regime
by a lapse of time during which the temperatures show a shows positive (warming) and negative (cooling)“spike-
reduced variation and maintain themselves at a constant like” fluctuations, that are sharp in the first and last winter,
value below 0 “C. This behavior can be attributed to the suggesting atmosphere-ground interaction. At the end of

Ribolini, A,: Active rock glacier in the southernmost permafrost environment., .
AT
('cclhr)
.b :":*
ST
ST
/
SGT
no data
. ..
...
. .
I .
, * '
ZlSep OlJan 071ar 19Jul 08.Nov Web 24Yay 234111
Figure 2. Hourly temperature change (DT, "Clhr) at DTL depth
every FR regime a rapid increase in SGT occurs, followed
by a close 0 "C temperature plateaux (SM). These sharp
temperature rises are decoupled from the daily air temperature
variations and could be referred to nonconductive
processes such as snow melt-water infiltration.
The 0 "C flat trace of several days, that follows every
rapid increase (Fig.l), could represent the ice-water phase
change. In order to evaluate the relations between the
presence of snow on the rock glacier surface and non
conductive processes, the ground-air thermal signal couplingldecoupling
could be considered. The magnitude of
ZC effect strongly depends on the soil water content, being
longer in response to enhanced moisture. The relation
between snow cover and the limitedlabsence of ZC for the
1998 and 2000 autumn suggests that the air temperature
lowering affected a well drained ground without a consistent
water content. During the 1999 and 2001 FR regimes,
the snow cover presence assured a relative water content
inside the block voids, triggering and prolonging the ZC
effect. The 1998-99 and 2000-01 winter recordings (FR)
clearly show coupled air-ground thermal traces, with
trends evidently disturbed by abrupt SGT. Funnels in the
thin snow cover likely allow the replacement of warm air
present in the block voids by dense cold air descending
from the atmosphere. Spike-like thermal anomalies are
absent or very limited during 1999-00 and 2000-01, because
the snow cover isolates the ground from external air
fluctuations. The warming magnitude relative to snowmelt
water infiltration in the frozen soil (SM) mainly results from
the infiltrated water volumeltemperature and soil porosity.
The coarse-grained ground surface of the Portette rock
glacier allows rapid infiltration of snow meltwater. To this
non-conductive process corresponds a high thermal impact
resulting in an up to 4-5 "C increase within a few
days. Finally, the long-lasting snow cover until late spring
insulates the ground from the incoming solar radiation and
prevents the ground from heating up.
The Portette rock glacier shows SGT trends characterized
by non-conductive and seasonally driven heat
transfer processes, i.e. release of latent heat during phase
change, snow melt infiltration and air circulation. The
ground-air temperature comparison suggests that the
permafrost growthldegradation strongly depends on the
timing of snow appearanceldisappearance at the rock glacier
surface. Delayed and thin snow cover in autumn represents
a forcing factor for permafrost maintenancelgrowing
because it allows i ) strong penetration of
cold into the ground, ii) lackinglreducing of the ZC effect
and ii) downward funnelling of cold air.
REFERENCES
Fabre, 0. and Ribolini, A. 2003. Limite actuelle des glaciers ro-chers
actifs dans le massif de I'Argentera (Alpes Maritimes italiennes).
Reunion du Societe Hydrotechnique de France, March 2003,
Grenoble.
Kane, D.L., Hinkel, K. M., Goering, D.J., Hinzman L.D. and Outcalt,
S.I. 2001. Non-conductive heat transfer associated with frozen
soil. Global and Planetary Change 29: 275-292.
134

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Undercooled scree slopes covered with stunted
abiotic factors to characterize the
dwarf trees in Switzerland -
phenomenon
A. Risf, M. Phillips and K. Auerswald*
Swiss Federal Institute for Snow and Avalanche Research (SLF), Davos Doe Switzerland
*Chair of Grassland Sciences, Department of Plant Sciences, Technical University of Munich- Weihenstephan, Freising,
Germany
INTRODUCTION
Large-scale permafrost distribution can be estimated using roundings and negative if azonal permafrost is to occur. The
numerical models simulating the spatial distribution of penna- energy balance at a site is determined by four energy balfrost
in complex topography (Hoelzle et al. 2001). The phe ance factors: radiation balance, ground heat flux, sensitive
nornenon can exist sporadically far below the alpine timber- heat flux and latent energy flux.
line as azonal permafrost (Haeberli 1978) at special locations
with extreme shadow and a significant lower mean annual
ground temperature than their surroundings. Sites satisfying
this definition are called undercooled scree slopes by
Wakonigg (1996) including locations which are not frozen
perennially,butsignificantlylongerthantheirsurroundings.
Undercooled scree slopes are characterized physiognomically
by a coverage of stunted dwarf trees (Fig. I), whose
growth is limited by the cold (Sieghardt et 2000). al.
Already in the 19th century an air stream through the
voids of the coarse blocky material of the scree slope
changing its direction seasonally was regarded as the causal
process for the exceptionally cold temperatures. This can be
explained by the relatively constant temperature of the air in
the underground compared to the seasonal temperature fluctuations
in the above-ground atmosphere: in summer the relatively
cold air inside the scree slope sinks downwards
through the big voids between the blocks and thereby draws
warm atmospheric air through the upper openings of the scree
slope. On its way through the underground the air cools
down on contact with the cold blocks and comes out as an
icy wind at the slope foot. In winter, when the rock is relatively
warm, the air stream changes its direction: cold atrnospheric
air streams into the lower openings of the scree slope,
warms on the blocks, which thereby cool down, and ascends
due to its lower density in order to escape into the atmosphere
at the upper end of the scree slope. This theory
only explains why the ground temperatures of an underigu~~
1. ~ n d ~ ~ s~ree ~ ~ slop ~ l~r~~~ e d du Van in the ~wis~ ~~r~
cooled scree slope are significantly lower than the surroundrnou~~a~n~.
The stunte~ ~ w t~ee~ a ~ (white f r ~ i~~i€ate m ~ ~ ~ u~de
ing air temperature in summer.
€QQi~~ ~ ~ n d i ~ i ~ n ~ .
However, out of this it can not be derived that the mean
annual ground temperature of this site is lower than in the surroundings,
because the lack of energy in winter might be
compensated by the surplus of energy in summer. The presented
theory can only be kept if the winter effect preponderates
the summer effect. Therefore an asymmetry in the sys-
tem has to be looked for. However, the energy balance of an
undercooled scree slope has to be lower than in its sur-
Our objective was to determine the abiotic factors of
dercooled scree slopes. In Switzerland 36 sites were investigated.
On the landscape scale, climatic, relief and ground
parameters were determined with GIs-techniques and com-
pared with the equivalent values of 495 reference sites with
zonal forest. Ten of the undercooled scree slopes were stud-
135
un

ied in more detail by comparing each with its reference site
with zonal forest in the immediate surroundings in respect b
climatic, relief, surface and ground parameters. For that purpose
each study slope was divided vertically into a high
(HZ), middle (MZ) and low zone (LZ) and horizontally into a
transect of stunted dwarf trees and one of zonal forest at the
slope foot (Fig. 2).
Rist, A. et al.: Undercooled scree slopes covered with stunted dwarf trees..
the lower energy level will be conserved by thermal insulation.
In spring this can result in a delayed snowmelt, Le. the
undercooled scree slope will become snowfree later than the
surroundings. Due to the higher short wave albedof snow in
comparison with soil and vegetation this will further decrease
the annual areal radiation balance of the undercooled scree
slope as compared to its surroundings.
Most of the investigated parameters differ only slightly
between an undercooled scree slope and its surroundings
and therefore the differences of the energy balances are also
marginal. Nevertheless, the physiognomical forms of forest d
the compared sites are different. Thus undercooled scree
slopes are labile systems and therefore represent a rare phe
nomenon and an ecosystem worthy of protection.
Figure 2. Zoning of an investigated slope horizontally in functional
transects, vertically in a high (HZ), middle (MZ), low zone (LZ) and
the rock face.
The investigated undercooled scree slopes are located in
areas with a mean annual precipitation between 1700 and
2000mm and at a mean altitude of over1000m a.s.1. The Osterreich 37.
presence of undercooled scree slopes is therefore restricted b
Wakonigg, H. 1996. Unterkuhlte Schutthalden. Arbeiten aus dem
lnstitut fur Geographie der Karl-Franzens-Universitat Graz Beimountain
regions, but not to an exceptionally cold zonal cli-
trage zur Permafrostforschung 33: 209-223.
mate. Undercooled scree slopes are characterized by special
local factors: the slope and the rock face above (present in
90% of cases) have a northerly aspect, are shady, and the
slope inclination is at least 20". The rock face - if present - is
at least 75m high and 40" steep. These relief parameters
cause a low annual radiation balance of about 2.3 GJlm2 on
average over the slope profile. In the transect of the stunted
dwarf trees respectively the undercooled scree slope the
surface area of open scree (about 40%) is more than twice
as high as in the proximate transect with zonal forest. There
fore the absorption of radiation will be less in the first transect,
as the short wave albedo of rock is higher than that of soil
and vegetation. Furthermore the air convection flux through
the coarse blocky material, which is important for the low energy
balance of an undercooled scree slope (Delaloye and
Reynard 2001), can be ascribed to the large surface area af
open scree and to the wide pore diameters (increasing from 4
cm on average at the head of the scree slope to IOcm at its
foot). The soil depth of the undercooled scree slope (8 cm on
average) is less by about one third than in the surrounding
zonal forest. The outcome of this is a reduced thermal insulation
of the undercooled scree slope. In combination with the
stronger air convection this may lead to a faster cooling of the
undercooled scree slope in comparison to the surrounding
forest in autumn. When the snow cover reaches about I m,
REFERENCES
Delaloye, R. and Reynard, E. 2001. Les eboulis geles du Creux du
Van (Chaine du Jura, Suisse). Environnement Pkriglaciaires.
Notes et Comptes-Rendus du Groupe Regionalisation du Periglaciaires
8: l 18-129.
Haeberli, W. 1978. Special aspects of high mountain permafrost
methodology and zonation in the Alps. Proceedings of the Third
International Conference on Permafrost, NRC Ottawa, 1, 379-
384.
Hoelzle, M. Mittaz, C. Etzelmuller, B. and Haeberli, W. 2001. Surface
energy fluxes and distribution models relating to permafrost in
European Mountain Permafrost areas: an overview of current
developments. Permafrost and Periglacial Processes 12: 53-68.
Sieghardt, H., Punz, W. and Reimitz, R. 2000. Vergleichendanatomische
Studien an Pflanzen der Eppaner "Eislocher".
Verhandlungen der Zoologisch-Botanischen Gesellschaft in
136

Formation and evolution of the Shalbatana Valley System, Mars: A case
study for the interaction between magmatism and permafrost
J. A. P. Rodriguez, S. Sasaki, *R. 0. Kuzmin and ** R. Greeley
Department of Earth and Planetary Sciences, University of Tokyo, Japan
*Vernadsky Institute of Russian Academy of Sciences, Moscow, Russia
**Department of Geological Sciences, Arizona State University, Arizona, U.S.A
INTRODUCTION
A large concentration of seemingly collapsed features and
associated oufflow channels are found between western Lunae
Planum and western Arabia Term in the cratered highlands
(units Npll and Np12) and the volcanic plains (unit Hr).
Comparisons with terrestrial flood channels, such as those d
Channeled Scablands in eastem Washington, suggest that
the Martian channels were eroded by catastrophic floods.
Most of the Martian channels originate from chaotic terrains
interpreted to represent areas where the ground collapsed as
water (confined under the permafrost) was released under
high hydrostatic pressures within artesian basins. Understanding
the geologic history of the Shalbatana Valley System
(SVS) can improve our knowledge of groundwater recharging
and extraction mechanisms, and the derivation d
the history of valley formation in the highlands. Shalbatana
Vallis was interpreted as an oufflow channel excavated by
water flow from an ice-covered paleolake in Ganges
Chasma (Nedell et al. 1987) and from catastrophically releases
from confined aquifers as the permafrost seal (Clifford,
1987) was disrupted at three locations, forming chaotic regions
(Cabrol et al. 1997). Rodriguez et al. (2003) proposed
that the excavation of SVS also involved water released
from an extensive underground cavern system produced by
magma and permafrost interaction. They also proposed an
alternative hypothesis for the origin for the Shalbatana up
stream chaotic region; they suggest that a highly degraded
late Hesperian impact crater (SE part) collapsed over the
putative cavernous system. In this work, we have used 128
pixelsldegree MOLA DEMs to derive alternative hypotheses
for the origin the Basin-A and its chaotic material, as well as
for the origin of the flanking Valley system Ill (Fig. IA).
THE BASIN A
The abrupt widening of the main Shalbatana Valley occurs
at the junction between VS-I and B-A (Fig. 1 B). This is
also the transition of the main valley from incision into the
Noachian plateau (bounded by unit Np12 to the east) and b
the West by VSlll (Hesperian highland unit Hr). The chaotic
terrain in 6-A was used as evidence that this basin resulted
from collapse over an aquifer (Cabrol et al. 1997). The east
margin of the main SVS valley shows a change in orientation
where it bounds B-A, which resemble intercepted craters.
The topographic region TLI partly bounds the basin at an ele
vation of about 700 meters above TU and TL3 (Fig. 16). A
channel (Chn) extends from VSI into B-A, the margins of
which are partly bounded by TLI-A and TLI-B. This suggests
that the topography was continuous prior to channel
formation. The topographic and roughness characteristics d
these terrains resemble those of Crater A (Fig. I6 and Fig.
2). Moreover, unit TLI-C occurs along the flanks of the arcuate
cliff section. These observations are consistent with this
unit representing the floor of collapsed craters intercepted by
SVS. The karst-like features (seen on the top edges of B-A),
are consistent with collapse processes for this part of
the
valley and suggest that the permafrost might have been relatively
shallow in the past andlor there are efficient mechanisms
of vertical transport of volatiles from a deeper permafrost
region, such as fractures produced by surface extension
over intrusive nonemergent dikes. Structural control (Cabrol
et al. 1997) over the main valley determinedthedeepand
narrowU-shaped VSI, which is incised into the Noachian
highlands. Structural control along the eastern flank of the
main valley is suggested by the orientation, depth and slope
characteristics and we propose that the abrupt widening of the
main valley resulted from a change in lithologies. This hypothesis
is consistent with the distribution of the chaotic terrain,which
is rougher, more abundant, and topographically
higher on the flanks of VSIII. This suggests that the chaotic
material was shed from the VSlll upland. The position of the
deepest and smoothest topographic unit (TL3) is consistent
with its being a depositional area for material carried by the
channel (Chn). We propose that flow from VS-I intercepted a
series of collapsed craters in this region and possibly formed
a crater-lake system, which might have extended as far as
the end of VS-II. The geological materials composing the
western margin of this lake system were apparently volatile
rich and suggest that the pemafrost at a regional scale was/
137
is unequally distributed andlor has undergone different
logical processes.
gee

THE VSlll COMPLEX
Rodriguez, J.A.P. el A+: Formation and evolution of Elre Shalbatana Valley System ...
It has been proposed (Cabrol et al. 1997) that water from
the B-A region successively drained and excavated the dl
and d2 channels in the Hesperian plateau (Fig. 2). These two
channels were abandoned following the deepening of the
main channel towards the Simud valleys (Cabrol et ai.
1997). We have observed that this region forms a complex
system of valleys, which includes rimless quasi-circular de
pressions, elongate depressions, pit chains and deep U-
shaped valleys. The roughness of this region reveals a high
frequency, low amplitude component, which would be consistent
with extensive thermokarstic activity and other sub
surface degradation processes (Rodriguez et al. 2003; Kuzmin
et al. 2002).
The d2 system: The contact between the d2 system and
B-A forms a wide valley (Fig. 2, yellow dotted line), which
narrows and is incised by a relatively narrow, southdipping
central region (d2a). Channel d2a joins with d2b to the north
forming a deep U-shaped valley. B-A drainage by the dl and
d2 channel systems is consistent with the temporal ponding
of VS-1 , B-A andVS-11, which was subsequently lowered
as flow toward the Simud valley occurred (Kuzmin et al.
2002).
We propose that in a first stage flow took place from B-A
to the north, forming a drainage channel of which only the
northem d2b section remains. After drainage from B-A
ceased the northern margin of B-A became unstable and recession
occurred preferentially headward along the preexisting
drainage channel. Recession happened mainly as debris
flows, fed by material shed from the valley flanks to its central
region, flowed towards B-A, which resulted in valley
widening, excavation of the central channel d2a, and the
deposition of most of the chaotic unit TL2.
The dl system: This system forms an N-trending series
of depressions, which includes wide shallow regions (Fig. 2,
WSR), narrow enclosed-and elongate-regions (Fig. 2,
NEER), and deep, rimless quasicircular depressions (Fig. 2,
RQCD). This association does not seem to be the result d
surface flow and it is more consistent with subsidence related
to long-lasted subsurface degradation, possibly by water
flowing through underground conduits (Rodriguez et al. 2003).
Other features in the region consistent with large subsurface
processes include the floor of Crater A, which shows surface
modifications and an elongate depression that extends south
from the margin of crater A, forming a sequence of pits, which
shallow up and become narrower to finally disappear near the
margin of the VS 11. The presence of dike like ridges at the
bottom of grabens located within the VSlll complex are suggestive
of the observed featured to be genetically related ts
interactions between intrusive magmatism and permafrost
(Rodriguez et ai. 2003).
FINAL REMARKS
These regions of the SVS demonstrate the importance d
lithological and structural controls over the morphological
characteristics of highland valleys. Our observations suggest
that processes possibly related to debris flows and hermokarstic
degradation produced the knobby material in the
chaotic region within B-A. The observing evidence for extensive
underground denudation apparently had resulted in different
degrees of subsidence, which might have been related b
thedevelopment of cavernous systems. These issues are
important to understand the complexity of hydrological processes
involved in the formation of the SVS, to quantify the
amount of water involved, and the amount of material eroded
during its excavation and to stress the complexity and great
age range of the processes involved in the Martian perma-
REFERENCES
Nedell, S.S., Squyres, S.W. and Andersen, D.W. 1987. Origin and
evolution of the lavered deDoslts in the Valles Marineris. Mars.
lcarus 70: 409-441.
Clifford, S.M. 1993. A model for the hydrologic and climatic behavior
of water on Mars. Geoahvs. Res. Lett.. Vol. 98 No.E6. 19073-
I ,
11016.
Cabrol, N.A., Ednmon, A. G. and Dawidowicz, G. 1997. Model of outflow
generation by hydrothermal underpressure drainage in
volcano-tectonic environment, Shalbatana Vallis (Mars), Icarus,
125: 455-464.
Rodriguez, J.A.P., Sasaki, S. and Mi amoto, H. 2003. Nature and
Hydrological relevance of the Shaybatana complex underground
cavernous s stem ,Geophys. Res. Lett., Vol. 30 No.6, 1304, 10.
Kuzmin, R.O., ireeley, R. and Nelson, D. M. 2002. Mars: The morhological
evidences of late Amazonian water activity in Shalgatana
Vallls. Lunar and Planetary Science conference XXXIII.
Abstract 1087.
138

8th International Conference on Permafrost Extended Abstracts an Current Research and Newly Available ~~~~~~~~~~~
Distribution of ramparted ground-ice depressions, Llanpumsaint, southwest
Wales.
N. Ross, C. Harris, P. Brabham and *S. Campbell
School of Earth, Ocean and Planetary Sciences, Cardiff University, Cardiff, Wales
*Department of Earth Sciences, Countryside
Council for Wales, Bangor, Wales
INTRODUCTION
Geomorphological features recognised as ‘ramparted groundice
depressions’ are well represented in the Welsh landscape.
These features have been interpreted as some of the
finest examples of relict open-system pingos in the UK,
formedwithindiscontinuouspermafrostduringtheLate De
vensian or Loch Lomond Stadia1 (GS-1; Watson 1972).
Landforms include horseshoe-shaped, circular or complex
ramparts, enclosing water- andlor peat-filled basins.
Although they are distributed throughout Wales (Fig. l),
previous research indicates a remarkable density in the
southwest, particularly within the tributaries of the Afon Tefi
(e.g. Cledlyn Valley), northwards towards Aberystwyth
(Watson 1972).
Figure 1: Distribution of ramparted ground-ice depressions
Wales (Adapted from Ballantyne and Harris 1994).
in
The known distribution of ramparted depressions in Wales
is however, thought to represent only a proportion of the total
number of potential sites. Research aimed at establishing their
distribution, morphology, structure and origin has been initiated
to develop a framework for their conservation. This paper
concentrates on the spatial distribution and density d
these features, presenting preliminary results from a representative
location where ramparted depressions are documented,
but no rigorous investigation has previously been
conducted.
LOCATION AND RESULTS
Ramparted depressions were identified near Llanpumsaint by
Watson and Watson (1974) in aseries of papers on ramparted
depressions in southwest Wales during the 1970s.
Their research focused predominantly on the features of the
Cledlyn and Cletwr valleys however, and although other locations
were illustrated in maps in various publications, the
exact map references of these sites were never published.
Re-identification and geo-referencing of these “dots on maps”
is therefore a current research priority.
The ramparted depressions of Llanpumsaint are located
within an area of approximately km2 20 in the tributaries of the
Afon Gwili at altitudes between 90-150 m. Aerial photography
has enabled the identification of numerous impressive landforms
with ramparts enclosing depressions and impounding
marshland. In excess of I00 clearly defined ramparts are
identifiable, not including areas of complex micro-topography
that may also represent ramparted depressions. A particularly
impressive subset of features is located to the south of Llanpumsaint
at SN 420275.
Preliminary mapping of this locality indicates a remarkable
density (45 perkm2)of identiable ramparts (Fig. 2),
representing approximately 25% of the total number of identifiable
ramparted depressions proximal to
139

Ross: N. et ai.: Distribution of ramparted ground-ice depressions ...
Llanpumsaint. The relationship between the distribution d graphical setting of the Llanpumsaint ramparted depressions
these features and water courses is of note, with well- with those of the Cledlyn valley means that a similar mechadeveloped
forms located in the areas adjacent or to, between, nism of formation can be tentatively inferred.
channels. Without the support of field verification however, the Tne assignment of all ramparted depressions recorded in
impact of channel modification for agricultural purposes cannot Wales to the Loch Lomond Stadia1 (ES-I) period must be
be assessed, and the current network of water courses may questioned. Ground-ice mounds producing ramparted depresnot
reflect the natural state. Like the features of the Cledlyn sions of this number, size and density are unlikely to have
valley, those at Llanpumsaint are clustered on valley bottoms formed within 1 ka.
and on low gradient slopes. This common morphology and To test these initial hypotheses, more detailed investigageographical
setting suggests a common mechanism of for- tions are planned to:
mation.
i. produce a complete inventory of ramparted depression
localities in Wales using aerial photography;
WIDER IMPLICATIONS AND FURTHER RESEARCH
ii. define the extent of ramparted depressions within
identified sites, and assess their relative conservation
status;
Preliminary fieldwork and a review of aerial photography reiii.
investigate selected sites utilising a suite of geo.
sources confirm the initial assessment that the distribution d
physical techniques and vibrocoring to determine inramparted
depressions in Wales is much more extensive and
of a higher density than previously recorded. This has imternal
structure and stratigraphy.
portant implications for the genetic origin and age of these
features.
The context necessary for the formation of contemporary REFERENCES
ground-ice mounds depends not only on climate, but also on
Ballantyne, C.K. and Harris, C. 1994. The Periglaciation of Great
an array of geological and hydro-geological factors. The re
Britain. Cambridge University Press, Cambridge.
markablenumberanddensityoframparteddepressions in Gurney, S.D. 1995. A reassessment of the relict Pleistocene “pinsouthwest
Wales is therefore indicative of a combination d gos” of west Wales: hydraulic pingos or mineral palsas Quafactors
that facilitated their development. Potentially, these in- ternary Newsletter 77: 6-16.
clude the interaction of i) topography, ii) superficial geology, Watson, E. 1972. Pingos of Cardiganshire and the Latest Ice Limit.
Nature 236: 343-344.
iii) structural geology, iv) the extent and influence of the Late
Watson, E. and Watson, S. 1974. Remains of pingos in the Cletwr
Devensian ice sheet, v) subsurface hydrology, and vi) gm basin, south-west Wales. Geografiska Annaler 56A: 213-225.
thermal flux.
Recently, a hypothesis that the ramparted depressions d
the Cledlyn and Cletwr valleys may represent the relict
forms of mineral palsas has been propagated in the literature
(Gurney 1995). In our opinion the origin of these features remains
an open question. Investigation of their internal structures
remains critical in addressing this issue, together with
the location, morphology, volume of the ramparts, hydrological
setting and distribution of sediments susceptible to ice segregation.
The apparent similarities of morphology and g e
140

Groundwater under deep permafrost conditions
T. Ruskeeniemi, LAhonen, M. Paananen, R. Blomqvisf, P. Degnan, ** S. K. Frape, *** Mensen, **** K.Lehto, 1. Wiksfrom,
***** L. Moren, 1. Puigdomenech and ****** M. Snellman
Geological Survey of Finland, Espoo, Finland; * UK Nirex Ltd. , UK; ** University of waterloo, Waterloo, Canada
*** Ontario Power Generation, Toronto, Canada; **** Posiva Oy, Olkiluoto, Finland
***** Svensk Karnbranslehantering AB (SKB), Stocholm, Sweden
****** Consulting Engineers Saanio & Riekkola, Helsinki, Finland
INTRODUCTION
According to the geologic disposal concept in Scandinavia,
spent nuclear fuel will be sealed in corrosion-resistant
and watertight canisters, and then emplaced in holes
drilled at the bottom of tunnels excavated into the bedrock.
The canisters are to be surrounded with bentonite clay,
which expands when it absorbs water. In Scandinavia the
depth of any repository is to be approx 500 m below
ground surface. When the last canisters have been emplaced,
the tunnels are to be filled, e.g., with a mixture of
bentonite and crushed stone, and the shafts leading to the
repository closed.
In assessing the safety of the geologic disposal concept,
it is essential to assess what physical and chemical
conditions could evolve within the repository and surrounding
environment. The objective of performance and
safety assessments is to demonstrate the long-term safety
of a repository for long-lived radioactive waste and most
management agencies consider scenarios up to I my into
the future.
During at least the past 2 million years, repeated cycles
of glaciation have occurred in the Northern Hemisphere.
At present, a major part of the area to the north of
latitude 60"N is under permafrost Conditions, although
within Fennoscandia at the same latitudes, temperate
conditions prevail. Climate models based on astronomical
forcing predict a cold period after approximately 60 000
years. Cold, dry periods favour the formation of periglacial
conditions and the development of deep permafrost, such
that permafrost is expected to reoccur in northern and
central Europe and North America.
In order to predict the long-term performance of a geologic
repository the influence of periglacial and glacial
conditions must be considered. A key goal of this research
project is to enhance the scientific basis for safety and
performance assessment and to derive constrained permafrost
scenarios relevant to the evolution of crystalline
groundwater flow systems.
A number of issues were identified as possible targets
of study: 1) Freezing rate and depth extent of permafrost
in crystalline bedrock; 2) Cold saline water segregations
(cryopegs) and their role in transport processes, 3) Role of
non-frozen areas (taliks) as pathways for groundwater
flow, 4) Aggregation of solid methane hydrates (clathrates),
and 5) Effects of freezing on bedrock stability.
All these aspects are related to the long-term stability
of hydrogeological and hydrogeochemical conditions.
Some of these phenomena are reviewed at length in the
literature and much is known about their role in shallow
systems or in sedimentary environments. However, much
less has been published regarding cryogenic processes
and hydrogeology in crystalline rock under deep permafrost
conditions.
This PERMAFROST project is a co-operative international
undertaking jointly funded by the participants from
Finland (Geological Survey of Finland and Posiva), Sweden
(Svensk Karnbranslehantering, SKB), Great Britain
(UK Nirex Ltd) and Canada (Ontario Power Generation
and University of Waterloo). After a short feasibility phase,
the research at Lupin started in late 2001 (Ruskeeniemi et
al. 2002).
THE STUDY SITE
Although the occurrence of permafrost is widespread in
the Northern Hemisphere today, there are only a few locations
at which permafrost in crystalline terrain can be investigated.
Such is the case at the Lupin gold mine in Canadian
Arctic (Nunavut Territory), approximately 1300 km
north of Edmonton, and some 80 km south of the Arctic
Circle. The distance to the White Sea in the north is about
300 km.
The site is within the continuous permafrost zone. Today
the depth of the permafrost in the area extends to
depths between about 400 and 600 m. The temperature
measurements conducted in the mine indicate that permafrost
persists down to 541 m. It is assumed that much
of the deep permafrost in this part of Canada has been
formed during the Holocene, i.e., during the past 10,000
years.
The region is located in the continental sub-arctic climate
zone well above the timberline. The vegetation is
141

typical for tundra: grass and low-growing shrub species
and dwarf birch. The mean annual air temperature is
-1 1 "C, and the annual extreme values range from 49°C
to +31 "C. The annual mean precipitation is low, about 270
mm.
The Lupin gold deposit is located in an Archean
metaturbidite sequence. The formation of the ore is dated
at 2.5 Ga. The major lithological units at the site are crystalline
rock types, phyllite and quartz-feldspar gneiss.
However, their sedimentary counterparts mudstone and
graywackelquartzite are locally preserved and recognizable.
The gold ore is hosted by an amphibolitic iron formation.
RESEARCH AT LUPIN
To support an evaluation of the feasibility of the site and to
develop a preliminary conceptual geosphere model, the
research at Lupin began with the compilation of the available
historical information on the lithostratigraphy, structural
geology, hydrogeology, mine hydrology, Quaternary
geology, etc. The mine site investigation focused on the
collection of groundwater samples from the mine at existing
boreholes and drift-exposed discontinuities.
In general the mine is dry (mine discharge about 50 000
m3/a), with the majority of the inflow occurring from a few
long exploration boreholes and, to a lesser extent, from
certain fracture zones. One of the fracture zones is repeatedly
cut by the spiral mine access ramp, which provided
an opportunity to collect dripping water to the 680 m
level at which the seepage apparently drains. Water samples
were obtained from a vertical cross section between
the 87 m and 1 I30 m levels.
As suspected, all the samples from frozen levels turned
out to be contaminated by Na-CI brines used for drilling in
permafrost. In comparison, the boreholes at lower levels
provided insight into the chemistry of the deep groundwaters
below the permafrost. These waters are brackish to
saline, Na-CI dominant, with dissolved solids (TDS) ranging
between 2 to 29 glL and lacking all indication (high
S04, NO3 and tritium) of contamination.
One of the key activities at Lupin is to explore whether
the outfreezing process can result in the formation of a saline
groundwater layer at the base of the permafrost in
crystalline bedrock. An electromagnetic sounding carried
out in the area gave indications of a subhorizontal conductor
in the depth range of 400 to 700 m. It is assumed
that this conductor would mark the location of the base of
the permafrost. According to model calculations, a saline
water layer with TDS of 5-30glL could explain the observed
anomaly. Parallel to the field research, laboratory
freezing experiments were carried out at the University of
Waterloo, Canada with various groundwaters collected
from Canada and Scandinavia. In addition to detailed information
on the behaviour of major and trace elements
and isotopic fractionation during outfreezing, the experi-
ments demonstrated that it was difficult to generate significant
volumes of concentrated fluids from the small vol-
Ruskeeniemi. T, et al.: Groundwater under deep permafrost,,,
ume of groundwater available in crystalline bedrock. This
result appears to contradict the results from the geophysical
field study.
To resolve this dilemma, and to support other hydrogeochemical
studies, two 400 to 500 m long low-angle research
boreholes were drilled from mine adits in late 2002
in an attempt to intercept fractures close to the base of the
permafrost that might contain uncontaminated groundwater.
The preliminary results suggest that the groundwater
table has declined and unsaturated conditions may exist
below the permafrost. Similar information is determined
from hydraulic pressure head measurements in boreholes
on the 890 and 1130 m levels. Currently it is too early to
say whether this is typical for deep permafrost conditions
in an area with limited recharge and low hydraulic gradients,
or if it is simply the effect of 20 years of mine operation.
A small amount of groundwater is seeping into the
new boreholes from the overlying permafrost. Geochemical
and isotopic geochemical charaterization of these
samples is underway.
CONCLUSIONS
The overall goal of the studies at the Lupin mine is to
study permafrost features and processes, which could
have relevance in assessing the performance and safety
of a repository at depth. Thus information obtained must
be transferable from observations in the Lupin mine to a
generic nuclear waste repository in a similar geological
environment. The research at Lupin will continue to address
the hydrogeochemical questions, but major efforts
are also directed towards improving the understanding of
the hydrogeology below the permafrost, in particular, the
influence of taliks on groundwater flow conditions.
142
REFERENCES
Ruskeeniemi, T., Paananen, M., Ahonen, L., Kaija, J., Kuivamaki,
A,, Frape, S., Moren, L., and Degnan, P. 2002. Permafrost at
Lupin: Report of Phase I. Geological Survey of Finland, Nuclear
Waste Disposal Research. Report YST-112,59 p.

A new approach to dating firn accumulation in an ice cave in the Swiss
Jura mountains
F. Schlatfer, M. Sfoffel M. Monbaron and *M. luetscher
Department of Geosciences, Geography, University of Fribourg, Fribourg, Switzerland
*Swiss Institute for Speleology and Karstology (SISKA), La Chaux-de-Fonds, Switzerland
INTRODUCTION
Located at 1359 m a.s.1. on the south east side of the
Swiss Jura range (BiereND, 6”17’50”/ 46”33’47”), a collapsed
doline with a diameter of about 40 m acts as an
important source of snow accumulation during the winter
season. The particular morphology and the sufficient
depth (-45m) of the cavity enable the trapping of cold air.
This causes the freezing of infiltration water and the
transformation of the snow into ice. Over sixty ice caves
are recognized in the Swiss Jura. The St Livres ice cave is
the second largest in the range. In 1976, its volume was
estimated at 3,500 m3 (Gigon 1976). Due to the significant
melting of ice in the recent past, its volume is now reduced
to approximately one-third (Le. 1200 m3).
Next to its vertical pit, the cavity has a horizontal development
of 60 meters (Fig. l), ice flows along for about
40 m, leading to the lower part of the cavity. In this place,
an overhanging ice wall forms the front of this small glacier.
Thus it is possible to observe an almost continuous
outcrop of the glacier. The presence of organic matter in
the ice allows the identification of different layers. Moreover,
great number of branches and trunks are trapped in
the ice (Fig. 2). In several places, deformations of the ice
layers confirm the presence of significant flow of the ice
filling. This flux is also responsible for the transport of
woody material that fell down in the accumulation zone,
that is, tree trunks and branches slowly movingto the
deepest part of the cavity.
In the past, the front of the glacier was completely covered
by a big stalagmite of ice forming a dome. Freezing
water infiltration originating from the roof of the cavity
formed this ice. In the last ten years however, the ice cave
melted considerably and the ice dome totally disappeared.
Furthermore, the front of the glacier also retreated, liberating
trunks and branches. The ice contains essentially
four tree species, namely European beech (Fagus sylvatica
L,), Great maple (Acer pseudoplatanus L.), European
silver fir (Abies alba Mill.) and Norway spruce (Picea
abies (L.) Karst.). In contrast to the wood trapped in the
ice, only the two conifer species are abundant in the local
forest next to the ice cave.
OBJECTIVES
This study aims to date (relative age) all ice layers with
woody material. Dendrochronological methods (Schweingruber
1996) are used to cross-date the single layers
containing subfossil tree trunks and branches. Analysis of
living trees outside the ice cave will reveal the link between
the living and the subfossil samples, permitting an
absolute dating of the single layers. This work is expected
to provide important data on the process of firn accumulation
in ice caves.
METHODOLOGY
First, the front of the glacier in the ice cave is mapped in
detail. Special focus is placed on the identification of different
accumulation layers and the position of some 60
143

subfossil trunks and branches of Norway spruce (Picea
abies (L.) Karst.) trapped in the ice. Cross sections of
every tree are made using a chainsaw. In the laboratory,
the samples are dried and polished using sandpaper. In
order to determine the age and the yearly increment of the
samples, tree-ring widths are measured, detrended and
averaged (Schweingruber 1996). The resulting series of
the 60 sections are then crossdated in order to build a
floating chronology (relative age).
In a further step, living trees outside the ice cave are
sampled in order to obtain growth patterns of undisturbed
living trees in the neighbouring forests. From each tree, at
least two cores are extracted with increment borers. The
resulting reference curve covers approximately the last
200 years.
Finally, the absolute reference curve of living trees is
cross-dated with the floating chronology of trapped wood
inside the cave in order to close the gap between the
youngest subfossil and the oldest living trees.
Schlatter, F. et al.: A new approach to dating firn accumulation
CONCLUSION
Dendrochronological investigations in ice caves have only
rarely been attempted to date ice layers and the accumulation
of firn. Nevertheless, this approach will help to better
understand the processes at the origin of such ice filling
and will significantly improve our knowledge of these particular
permafrost occurrences, without distur-bing the
subterranean environment too much. Preliminary results
will be presented on the poster.
REFERENCES
Audetat M. and Heiss G. 2002. lnventaire speleologique du Jura vaudois,
partie ouest. Acadernie suisse des sciences naturelles, La
Chaux-de-Fonds: IRL Renens
Gigon R. 1976. lnventaire speleologique du canton de Neuchitel. Societe
helvktique des Sciences Naturelles. Neuchitel.
Schweingruber, F.H. 1996. Tree Rings and Environment. Dendroecology.
Swiss Federal Institute for Forest, Snow and Landscape
Research WSLIFNP. Birmensdotf:
144

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Rock glacier inventory of the Adamello Presanella massif (Central Alps,
Italy)
R. Seppi, *C. Baroni and **A. Carton
Dipartimento di Scienze della Terra, Universifa di Pavia, Pavia, lfaly and Museo Tridenfino di Scienze Naturali, Trenfo, Italy
*Dipartimento di Scienze della Terra, Universia di Pisa and CNR, lsfitufo di Geoscienze e Georisorse, Pisa, lfaly
**Dipartimento di Scienze della Terra, Universifh di Pavia, Pavia, Italy
The rock glacier inventory of the Adamello Presanella rock glaciers facing
N, NE and NW is considerably lower
Group (Central Alps, Italy) was carried out by means of than the elevation those facing S, SE and SW (Fig. 1). In
field and topographic surveys as well as aerial photograph the activelinactive forms this difference is 202 m, whereas
analysis. The study area is characterized by a series of in the relict ones it is 89 m.
valleys descending from the summit of the massif with a
centripetal pattern. Geomorphology is dominated by mass
N
wasting (talus slopes, debris cones, debris-flow cones and
lobes), and glacial (erosional landforms and moraines)
2700 m a.s.1.
and periglacial (mainly rock glaciers) landforms. Lithology
is characterized by the Adamello-Presanella-Monte Re di
Castello intrusive rock complex (tonalites and granodiorites).
The metamorphic basement and volcanic and
sedimentary rocks of the Permian-Cretaceous sequence
outcrop at the external margins of the batholite. The
mountain summit area hosts the largest glacier of the Italian
Alps (Adamello Glacier, 17.66 km2); more than 90 glaciers
are to be found in the northern and central sectors of
the massif.
In the study area 216 rock glaciers have been identified
and mapped by means of GIS (ArcView 3.2). They
have been divided into (a) activelinactive (83 rock glaciers,
38%) and (b) relict (133 rock glaciers, 62%), according
to Barsch (1996). The average elevation of the
front of activelinactive rock glaciers is 2493 m a.s.1. (mini-
S
mum value 2180 m a.s.1.). The average of the maximum
Lower limit activelinactive
elevations of the same rock glaciers is 2614 m a.s.1.
rock glacters
I I Lnwer limit relict
(maximum value 2940 m a.s.1.). The average elevation of
rock glaciers
the front of the relict forms is 2071 m a.s.1. (minimum
value 1700 m a.s.1,). The average maximum elevations of Figure 1, Altitude of the fronts of activelinactive and relict rock glaciers
the same rock glaciers is 2183 m a.s.1. (maximum value compared with aspect. The number of forms for each orientation is
2650 m a.s.1.). Activelinactive rock glaciers are on average shown.
291 m long and 163 m wide, whereas the relict ones are
on average some 290 rn long and I99 m wide. The average
area covered by rock glaciers is 0.041 km2, varying
from 0.039 km2 for the activelinactive forms and 0.043 km2
for the relict ones. The largest form (A75 in the inventory),
located in Val di Lares, is a lobate type and extends over
0.27 km2.
In terms of orientation, 61% of rock glaciers face N, NE
and NW, whereas 17% are exposed to the S, SEI SW.
The remaining ones face E (13%) and W (9%), respectively.
The elevation of the front of activelinactive and relict
Measurements of the displacement of rock blocks from
two active rock glaciers are in progress. Preliminary survey
has been carried out on each of them by means of a
theodolit. The position of 25 boulders marked with screw
anchors was checked one year after installation on a rock
glacier located in the upper Val di Genova (A37 in the in-
ventory). This tongue-shaped rock glacier, which ranges in
elevation from 2750 to 2880 m a.s.l., faces SW, is some
280 m long, 100 m wide and covers an area of about
0.023 km2. The movement of the single monitored boul-
145

ders ranges between 0.05 m and 0.21 m. The boulders
scattered on the frontal and left-sided portions of the rock
glacier (points 1-15) are subject to greater displacements,
whereas the higher sector is less dynamic (Fig. 2). In any
case, the measured movements are rather slow (cf.
Barsch 1996, Kaufmann 1996) and generally oriented
along the slope's maximum dip. The surface velocity surveyed
in the second rock glacier, which is located in Val
d'Amola (A42 in the inventory), ranges between 0.02 and
0.16 mlyear, with flow vectors oriented in a rather homogeneous
way. The eastern portion of the front appears to
be more dynamic (0.12-0.15 mlyear).
Seppi, R. et 81,: Ruck glacier inventory of the Adamello Presanelia massif.
riod. On the other hand, more forms (about 30) are placed
immediately outside the boundaries attained by glaciers
during the LIA and do not show any relationship with the
glacial deposits of this period. It may therefore be inferred
that among activelinactive rock glaciers the most recent
forms are ascribable to a period subsequent to the LIA,
whereas the other forms are attributable to coeval or preceding
Holocene phases.
REFERENCES
Barsch, D. 1996. Rock glaciers: indicators for the present and former
geoecology in high mountain environments. Berlin: Springer-
Verlag.
Frauenfelder, R., Haeberli, W., Hoelzle, M. and Maisch, M. 2001. Using
relict rock glaciers in GIS-based modelling to reconstruct
Younger Dryas permafrost distribution patterns in the Err-Julier
area, Swiss Alps. Norsk geog. Tidsskr. 55: 195-202.
Kaufmann, V. 1996. Geomorphometric monitoring of active rock glaciers
in the Austrian Alps. High Mountain Remote Sensing Cartography:
Proc. Intern. Symp., Karlstadt-Kiruna-Tromsra, 19-29
August 1996: 97-113.
Lambiel, C. and Reynard, E. 2001. Regional modelling of present,
past and future potential distribution of discontinuous permafrost
based on a rock glacier inventory in the Bagnes-Heremence area
(Western Swiss Alps). Norsk geog. Tidsskr. 55: 219-223.
Sailer, R. and Kerschner, H. 1999. Equilibriurn-line altitudes and rock
glaciers during the Younger Dryas cooling event, Ferwall group,
western Tyrol, Austria. Annals of Glaciology 28: 141-145.
Figure 2. Movement (direction and rate of displacement) of marked
boulders (1-25) on the active rock glacier located in Val di Genova
(A37 in the inventory). Sl=basis of topographic survey; S2=bench
mark of topographic survey.
The minimum altitude of activelinactive rock glaciers is
utilized as an indicator of the present lower boundary of
discontinuous permafrost (Barsch 1996), whereas the elevation
of relict forms is used as an indicator of past permafrost
conditions. In the study area the difference in altitude
between the fronts of the activelinactive and relict
forms indicates a increase in the lower boundary of discontinuous
permafrost of 420 m. In other alpine areas, the
relict rock glaciers have been ascribed to the Late Glacial
(Sailer and Kerschner 1999, Frauenfelder et al. 2001,
Lambiel and Reynard 2001). If the relict rock glaciers of
the Adamello Presanella massif were tentatively correlated
to Late Glacial by applying a mean temperature
lapse rate of 0.65 "CIIOO m to the altitude variation of the
lower limit of the discontinuous permafrost, a mean annual
temperature rise of about 2.7 "C would be recorded.
Among the activelinactive rock glaciers, about 20 are
located in areas occupied by glaciers during the Little Ice
Age (LIA) or are fed by glacial deposits from the same pe-
146

8th International Conference
Permafrost Extended Abstracts on Current Research and Newly Available Information
Viable protozoa from permafrost sediments and buried soils
A. V. Shatilovich, LAShmakova, S. V.Gubin and D.A. Gilichinsky
Soil Cryology Laboratory, Institute of Physicochemical and Biological Problems in Soil Science, Russian Academy
Pushchino, Moscow Region, Russia
of Sciences,
INTRODUCTION
A significant number of viable ancient microorganisms is
known to be present within permafrost. The age of the cells
corresponds to the longevity of the permanently frozen state
of the sediments, and the oldest date back to the Pliocene
(Gilichinsky 2002). Viable spores of the lower plants
(mosses) and seeds of the higher plants were also found
within Siberian and Canadian late Pleistocene permafrost; it
was shown that they were still able to grow (Porsild et al.
1967; Gilichinsky et al. 2001; Gubin et al. 2003). In order b
extend our knowledge of biotic survival ongeologicaltime
scales within permafrost, we aimed at the detection of viable
protozoa. Protozoa represent a considerable part of soil biota
and possess a wide spectrum of adaptation abilities.
MATERIALS AND METHODS
Figure 1. Geological section of icy-loess complex.
The late Pleistocene icy-loess complex, Le., syngenetically
frozen aleurites with the traces of soil (so-called cryopedolite), RESULTS AND DISCUSSION
ice veins and buried soils on Kolyma lowland, has been
studied. Cryopedolite contains up to 1% organic matter and Organic matter from burrows contain a large number of ele
thin filamentous roots of grasses. Buried soils belong to His- ments from late-Pleistocene tundra-steppe ecosystems. This
tosols and Gleysols. We also investigated samples from bur- great volume of paleoecological information survived due b
rows of fossil gopher. The studied material consisted of mix- the frozen state of deposits from the time of their formation. Viture
of aleurites with litter components: seeds, residues d able protozoa were found among other biological objects.
leaves and grasses, wool of animals, etc. (Fig. I). Radiocar- Protists were isolated from all samples but one (collected
bon age of burrows was determined as 28 to 32 thousand from the top layer of buried Histosol) and were identified as
years (Gubin et al. 2001).
representatives of three large taxones: heterotrophic flagel-
In order to find viable protozoa we used a method of soil lates, naked amoebae (Fig. 2), and small ciliates. Heterotroenrichmentcultivationonapoornutrientnon-selective
me phic flagellates were dominant species of recovered commu-
dium (sterile water and Lozina-Lozinskii medium). The cultivation
was carried out during one month using two different
temperatureAight conditions: 2OoC with daylight in one case
and 5°C without light in the other. Protozoa were observed
directly in the soil enrichment cultures and in suspension.
Methods of light and fluorescent microscopy were used
their morphological description and identification. images
for
d
protists were obtained by means of a digital camera.
nities. It was observed that a succession cycle (species
composition change) of permafrost protests community takes
approximately 20 days. A tendency appeared to exist for an
increase in the number of viable protists and their species diversity
in soil samples with high content of fossil plant . It
was also observed that protozoa abundance and biodiversity
was higher inside fossil burrows than in buried soils. This fad
is likely to be due to initial relative abundance of protozoan
fauna as well as to more favorable conditions of its cryopre-
147

servation within the burrows. Following long-term survival in
permafrost, the organisms become subjected to the stresses
of thawing and exposure to oxygen as the samples are
melted in the lab environment at relatively high temperatures.
Becausethestrategiesandtechniquesforthe recovery d
protozoa from frozen environments are just beginning to be
developed, the standard methods for isolation of the protists in
purecultureshavenotyetdeliveredpositiveresults. Most
protozoa have not been cultured, and many of them died OT
altered their activity.
ACKNOWLEDGEMENTS
This work was supported by the Russian Foundation of Basic
Research (grants 01-05-65043 and 01-0449087).
REFERENCES
Gilichinsky, 0. Shatilovich A,, Vishnivetskaya T.., Spirina E., Faizutdinova
R., Rivkina E., Gubin S., Erokhina L., Vorobyova E., Soina
V. 2001. How long the life might be preserved In (Giovannelli
F., ed.), Proceedings of the International Workshop ((The
Bridge between the Big Bang and Biology)), Consiglio Nazionale
delle Ricerche of Italy: 362-369.
Gilichinsky, D. 2002. Permafrost as a microbial habitat I/ Encyclopedia
of Environmental Microbiology, 2002, 932-956.
Gubin, S.V. et ai. 2001. Composition of seeds from fossil gopher
burrows in the ice-loess deposits of Zelyony mys asenvironmental
indicator. Earth Cryosphere 5(2): 76-82 (in Russian).
Gubin, S.V. et al. 2003. The Possible Contribution of Late Pleistocene
Biota to Biodiversity in Present Permafrost Zone /I Journal
of General Biology 2003, v.64 -1: 45-50 (in Russian).
Porsild, A. et al. 1967. Grown seeds of Pleistocene age /I Science, v.
158: 113-114.
a.
CONCLUSION
For the first time, viable protozoa were discovered in frozen
layers and their ability to survive within permafrost during
dozens of thousands of years was shown. This suggests that
in such a unique physicochemical environment as permafrost,
these organisms keep unknown possibilities of physiological
and biochemical adaptation incomparably longer than
in any other known habitat. Their ability to survive on gee
logical time scales within the broad limits of natural systems
forces us to redefine the spatial and temporal limits of the terrestrial
biosphere. On the basis of the obtained results, it is
possible to propose the existence of viable protozoa not only
in buried soils, but also in deposits rich in organic matter.
Permafrost thawing due to natural or anthropogenic processes
exposes ancient life to modem ecosystems and the ancient
protozoa renew their physiological activity and participate in
formation of modern biodiversity on the northern territories.
148

Stress-strain conditions and the stability of arctic coastal slopes:
assessments using seismic surveys
A. G. Skvortsov and D.S. Drozdov
Earth Cryosphere Institute, SB RAS, Russia
Within the framework of ACD and INTAS-2332 projects,
seismic research for the purpose of slope stability analysis
was carried out at the Bolvanskii key-site in the mouth of
Pechora river (Barents sea). A special seismic technique
for investigations at various types of slopes, mainly at
sliding slopes, was used. It is based on theoretical relationships
and experimental correlation between seismic
characteristics of the ground and stress-strain conditions
of the soil. A very important feature of this technique is the
possibility to monitor and predict preliminary changes of
slope stability before discontinuity occurs. The term of forecast
ranges from some months to several years. The
uncertainty in spatial allocation of expected discontinuities
does not exceed a few meters (Gorjainov et al. 1987,
Skvortsov 1989).
In previous years, the most adequate results were obtained
for different sites in temperate climatic zones. The
geological conditions of some examined sites can be considered
to be a model of Arctic frozen slopes in summer
time. The most interesting data was gained while investigating
a coastal slope where limestone underlies a layer
of clay. The processing of seismic data showed that the
stable (for the date of research) part of the slope is going
to be involved in the sliding process. Mainly the seismic
data on the upper horizon of the clay layer were used to
trace the position of the predicted crack (Alyoshin et al.
2002). Within a few years, new landslides took place according
to the forecast.
The results of this research and numerous other experiments
have shown that reliable information on the
stress-strain conditions in slopes can be received on the
basis of data on compressional wave velocity and its anisotropy
in the uppermost part of a geological section (1 m
thickness). From the methodical point of view this conclusion
is extremely important for assessments of arctic
coastal slope stability, since it shows that the stress-strain
conditions of the coastal massif can be examined using
the distribution of seismic characteristics in the active
layer (seasonally-thawed layer).
This conclusion was taken into account while carrying
out in-situ field investigations at the northern slope of cape
Bolvanskii where sea water undermines the third marine
terrace (September, 2002). Three main types of geological
cross-sections are observed along the coast: 1) mainly
peaty in top part; 2) mainly sandy; 3) mainly loamy. The
loams contain many boulders. All soils are permanently
frozen. Generally, they have relatively low ice-content (no
more than 0.2). The increased ice-content (up to 0.3-0.4)
is a characteristic of the upper 2-3 m of loamy crosssections.
The peat bogs are the most ice-rich sites which
include thick ice wedges (Malkova et al. 2002).
At the seismic key-site sand prevails in cross-section.
The height of the coastal slope is about 25 m and the
slope is rather abrupt (30"). However, any modern active
processes, including landslides, crumbling and slumping,
are not observed now while at nearby sites all these processes
occur. Dense bushes, as seen on Figure 1, prove
the modern stability of the investigated slope.
The seismic operations were executed at four 50-60 m
long profiles located perpendicularly to the coastal line.
The profiles cross a sandy surface stretched along a ledge
(bright grey band on Fig. I) and enter a peatbog (grey
color). Information about the distribution of compressional
and shear wave velocity, both in the active layer and in the
top part of permanently frozen ground, was obtained. The
preliminarily processing of these results showed obviously
that the wave velocity greatly depends not only on the
stress-strain distribution and further expected disturbances,
but also on the variability of lithology, moisture,
density, ice and organic content, etc.
Other geophysical parameters depend less on these
factors: the lateral anisotropy of wave velocity, Poisson's
ratio and its anisotropy. To obtain them the wave velocity
measurements were carried out in lateral rectangular directions
(Skvortsov and Drozdov 2002).
The analysis of seismic characteristics of the seasonally
thawed ground allows to define two weakened zones
in the active layer. Most obvious, these zones are seen at
the isoline maps of compressional wave velocity anisotropy
and of Poisson's ratio anisotropy (Fig. 2).
The first more sharply expressed weakened zone is
situated in the north-western part of the site. It is subparallel
to the ledge of terrace and is indicated as a zone
with minimum values of wave velocity anisotropy and
Poisson's ratio anisotropy.
149

~igu~e 1. ~ ~ ~rQfiles ~ at the olvanskii s ~ ke~-sit~ i ~
the ~ ~ ~ a~ i a n~ e ~ r~
a o ~ ~ ~ ~ i n ~ .
Skvortsov, A. G. et al.: Stress-strain conditions and the stability of ...
and the ~ e~ult~ of
An attempt has been undertaken to detect the presence
of landscape attributes of this zone at the surface.
For this purpose micro-landscape mapping was carried
out during which an existing crack in the ground surface
has been found 20 m to the west of the key-site in the first
zone. That proves the first zone to be a weakened one,
The first weakened zone indicates a line along which the
separation of relatively small frozen blocks can move
(“expected crack at Figures 1 and 2).
The second weakened zone shows less contrast on
isoline maps than the first zone. It is situated at 20-40
meters from the slope, crosses the first zone and ledge
and probably is an attribute of larger frozen blocks which
can be separated in a comparatively longer future. (“anisotropic
ground strain” on Fig. I and 2).
The submitted results illustrate a possibility of surface
seismic techniques to investigate slope processes, to assess
the stability of frozen ground and make spatio - tem-
poral forecasts of coastal retreat at particular sites in the
Arctic.
ACKNOWLEDGEMENTS
This project is supported by INTAS (grant 01-2332) and
SB RAS grant.
The authors express gratitude for cooperation and assistance
with carrying out in-situ field investigations to:
Norwegian Geotechnical Institute (manager of Oil Pollu-
tion project 0. Nerland) and Circumpolar Active Layer
Monitoring project (CALM, manager on European Russian
sites G.V. Malkova).
REFERENCES
Alyoshin, AS., Drozdov, D.S., Dubovskoj, V.B. and Skvortsov, A.G.
2002. New opportunities of geophysical monitoring of processes
slope. Sergeev reading, issue 4. - Moscow; GEOS: 485-488 (in
Russian).
Gorjainov N.N., Bogolyubov, N.M., Varlamov, A.N., Matveev, V.S.,
Nikitin, V.N. and Skvortsov, A.G. 1987. Studying of landslides by
geophysical methods - Moscow; Nedra: 157 (in Russian).
Malkova (Ananjeva), G.V., Drozdov, D.S., Kanevskiy, M.Z. and
Korostelev, Yu.V. 2002. The coastal researchs in the area of
geocryological station “Cape Bolvanskiy ‘I, the estuary of Pechora
river. 3-rd Workshop on Arctic Coastal Dynamics. - Oslo, IPA:
27-28.
Skvortsov, A.G. 1989. Monitoring of sliding slopes stability by use of
seismic methods. Investig. of hydro-geol. and engineer.-geol. objects
by geophysical and isotope methods. - Moscow: VSEG-
INGEO: 3243 (in Russian).
Skvortsov, A.G. and Drozdov, D.S. 2002. The assessment of stressstrain
conditions of coastal slope by use of seismoreconnaissance
(seismic-service). 3-rd Workshop an Arctic Coastal Dynamics. -
Oslo. IPA: 48.
150

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Contraction crack dynamics in polygonal patterned ground and soil
inflation in the Dry Valleys, Antarctica
R.S. Sleften and B. Hallet
Quaternary Research Center, University of Washington Seattle, WA, USA
INTRODUCTION
The Dry Valleys of Antarctica are believed to be part of an
ancient landscape that has remained undisturbed for millions
of years. Our recent work suggests, however, that
the soils and land surfaces are relatively active due in part
to seasonal opening and partial closure of contraction
cracks that delineate the pervasive polygonal patterned
ground in this region (PBwB 1959). Unlike the relatively
moist Arctic where contraction cracks tend to fill with surface
meltwater and form ice wedges, contraction cracks in
the Dry Valleys tend to fill with sand and form sand
wedges. The growth of sand wedges disturbs the ground
surface and the development of soil profile. We examined
the rates of sand wedge growth by continuing a four decade-long
survey of displacements across contraction
cracks, and by installing modern instrumentation.
RESULTS AND DISCUSSION
To measure the growth of sand wedges, Robert F. Black
and Thomas Berg (Berg and Black 1966; Black 1973) installed
over 500 pairs of 60cm vertical rods that span contraction
cracks in the Dry Valleys during the early to mid
1960s. They measured crack divergence manually using
precision rulers. We revisited and measured Black’s 13
sites on three occasions since 1995; we found the rods to
be generally in good condition and that crack divergence
continued at rates (0.4-2 mm a”) similar to those measured
by Black (Figure 1).
To provide a better understanding and a basis for
modeling the annual dynamics of contraction cracks, we
installed linear transducers connected to data loggers at 3
sites in Beacon Valley and Victoria Valley in 1998, and
have continuously monitored displacements across cracks
since that time. In addition, we monitor the micrometerology,
soil temperature, soil thermal properties, and other
physical parameters.
The total annual crack opening due to thermal contraction
of polygons ranges from 0.5 cm to nearly 2 cm at
other sites, and lags air temperature (figure 2). The con-
30 I
25
20
15
10
5
0
n
0 IO 20 30 40
-3-
3 Time since (years) installation
v
Figure 1, Average contraction crack growth as determined by rods installed
at 3 Black’s sites and measured manually. Initial and 7-year
measurements were made by Black. Our measurements were made
35 and 39 years after the rods were pounded into the permafrost.
traction cracks open during cooling as the ice-rich ground
contracts and expands as the ground warms. The net annual
crack divergence reflects the amount of fines that fall
into the cracks thereby preventing full closure. The 40-
year average rate of divergence across the cracks based
on Black‘s rods is 0.6 mm a-1 for the lower Beacon Valley
site, whereas our continuous measurements for the past 4
years show a net crack divergence rate of 1.4 mm a”. If
the more rapid recent growth is significant, it presumably
reflects an increase in either the amplitude of the thermal
forcing or the sediment input into cracks.
The long-term growth of sand wedges has caused an
effective inflation of the soils and, in lower Beacon Valley,
raised the ground surface by at least 8 m, the depth of our
deepest borehole there. Preferential injection of finegrained
material to depth, as the larger rocks do not fit into
the open contraction cracks, ultimately results in the accumulation
of rocks and boulders on the surface. The
sand wedges grow to encompass the entire surface and
appear to repeatedly sweep across the surface, thereby
building up a complex of sand wedges overlain by rocks.

Sletten, R.S. et al.: Contraction crack dynamics in polygonal patterned ground
REFERENCES
20
0
-20
-40
3
8 -60
6 16
3
h
14
".
Y
12
10
8
6
4
0
$ 2
d
% - O
8 1199 7199 1/00 7100 1/01 7101 1102 7102 1103
"-.
3 Date
Y
Figure 2. Continuous record of surface temperature and divergence
across contraction cracks at the periphery of sand-wedge polygons
using linear transducers mounted on rods spanning the cracks. The
dashed line indicates a net annual divergence rate of 1.4 mm a.1.
Berg, T. E. and Black, R. F. 1966. Preliminary measurements of
growth of nonsorted polygons, Victoria Land, Antarctica. Antarctic
Soils and Soil Forming Processes, Antarctic Research Series, Vol.
8. J. C. F. Tedrow. American Geophysical Union, Washington,
DC: 61-108.
Black, R. F. 1973. Growth of patterned ground in Victoria Land, Antarctica.
Permafrost: The North American Contribution to the 2nd
International Conference, Yakutsk, Siberia, National Academy of
Sciences: 193-203.
Schirrmeister, L., Sieged, C. Kuznetsova, T. Kuzmina, S. Andreev, A.
Kienast, F. Meyer, H. and Bobrov, A. 2002. Paleoenvironmental
and paleoclimatic records from permafrost deposits in the Arctic
region of Northern Siberia. Quaternary International 89(1): 97-118.
Pewe, T. L. 1959. Sand-wedge polygons (Tesselations) in the
McMurdo Sound region, Antarctica-a progress report. American
Journal of Science 257: 545-552.
This injection of fine-grained sediment into contraction
cracks and the resulting inflation of the surface in the Dry
Valleys appear analogous to the formation of the ice complexes
in Siberia due to the recurrent build-up of ice in
contraction cracks (Schirrmeister et al. 2002). In both
cases, soil inflation has been occurring for long periods.
Our preliminary cosmogenic isotope data indicate that
sand from sand wedges at 5 m depth has been buried for
several million years, suggesting long-term burial and inflation
and, possibly, shielding by the Taylor Glacier during
a former excursion into Beacon Valley.
CONCLUSIONS
Current rates of sand-wedge growth (-1 mm a") and
polygon sizes (-10 m) suggest that the near-surface layer
of polygons would be completely resurfaced, in a period of
tens of thousands of years, while rocks remain at the surface
indefinitely as sand-wedges grow beneath them. The
continuous injection of fines into contraction cracks has
created a sand wedge complex that has inflated the surface
over 8m. Our findings suggest that landscape surfaces
with well-developed polygonal patterned ground in
the Dry Valleys are relatively active, which contrasts with
suggestions that these surfaces have been undisturbed
for millions of years.
152

The Global Terrestrial Network for Permafrost (GTN-P) - status and
preliminary results of the thermal monitoring component
S.L. Smith, M. M. Burgess, *V. Romanovsky and **J. Brown
Geological Survey of Canada, Natural Resources Canada
*University of Alaska, Fairbanks, USA
**International Permafrost Association, Woods Hole, Mass., USA
INTRODUCTION
The Global Terrestrial Network for Permafrost (GTN-P) was
established in 1999 to provide long-term field observations of
active layer and permafrost thermal state that are required b
determine the present permafrost conditions and to detect
changes in permafrost stability. The data supplied by this
network will enhance our ability to predict the consequences
of permafrost degradation and to develop adaptation measures
to respond to these changes. The GTN-P contributes to the
World Meteorological Organization's Global Climate Observing
System and Global Terrestrial Observing System.
An overview of the GTN-P is given by Burgess et al.
(2000).
The active layer component of the GTN-PI the Circumpolar
Active Layer Monitoring (CALM) program, has been in
operation over the last decade. Organizational efforts of the
GTN-P are therefore focused on the recently established
borehole thermal monitoring component. This paper describes
the present status of the thermal monitoring component and
presents preliminary results from North America.
STATUS OF THE THERMAL MONITORING
PROGRAM
The permafrost thermal monitoring program consists of a
globally comprehensive network of boreholes for ground temperature
measurements. Many of these boreholes were
drilled for research, geotechnical, or resource exploration purposes
in the last two decades and have been maintained as
thermal monitoring sites. About 370 boreholes from 16 countries
have been identified as candidate sites for inclusion in
the GTN-P thermal monitoring system (Fig. I). The majority
of these boreholes are between 10 and 125 meters deep and
areinthenorthemhemisphere.A few sitesarelocatedin
Antarctica and Argentina. Regional networks include those of
the Geological Survey of Canada (GSC) in the Mackenzie
region, the University of Alaska's Alaskan transect, the
United States Geological Survey's deep boreholes in Northem
Alaska and the European Community's Permafrost and
Climate in Europe (PACE) program of boreholes largely in
alpine permafrost. A web site (http://sts.qsc,nrcan.qc.ca/atnp)
developed by the GSC, provides an inventory of candidate
boreholes, and background material on the GTN-P.
Metadata compilation for the candidate boreholes has
been conducted over the last two years. Metadata from about
60% of the boreholes has been compiled and posted on the
ETN-P website. Over the next year, metadatacompilation
will be completed and site selection will take place. Compilation
of summary historical data for network sites will also occur
over the next year.
PRELIMINARY RESULTS
A general discussion of preliminary results from GTN-P sites
canbefound in Romanovsky et al. (2002). Thispaper b
cuses on results from across the North American Arctic.
Analysis of permafrost temperatures, taken in theupper
20 m, from a network of boreholes in Alaska maintained by
the University of Alaska Fairbanks indicates that permafrost
has warmed over the last 20 to 50 years. At the recently reactivated
monitoring site at Barrow, temperatures at a depth d
15 m increased 1°C between 1950 and 2001. Increases of
permafrost temperature ranging from 0.6 to 1.5"C between
1983 and 2000 at a depth of 20 m have been observed at
sites along the Trans-Alaska pipeline route.
The GSC has operated a network of thermalmonitoring
sites in the Mackenzie region, Northwest Territories since
1985. At 15 m depth at a site in the central Mackenzie valley
an increase of about 0.03"C per year was observed be
tween 1986 and 2001 in permafrost with temperatures of ap
proximately -1°C. In the northem Mackenzie region, in
colder permafrost (-7"C), temperatures at a depth of 28 m
show an increase of about 0.1 "C per year in the 1990s.
At the GSC's high Arctic observatory at Alert, a warming
of the permafrost occurred in the late 1990s of 0.15°C per
year at a depth of 15 m. Some cooling of permafrost was
observed from the late 1980s to the early 1990s in the eastern
Canadian Arctic at a depth of 5 m at Environment Canada's
borehole at Iqaluit. This cooling however, was following by
warming in the late 1990s. This trend is similar to that
observed in Northern Quebec, where cooling of about 0.1 "C
153

Smith, S.L. et al.: The Global Terrestrial Network for Permafrost (GTN-P) ...
Figure 1. Permafrost distribution in the Northern Hemisphere and location of candidate boreholes for permafrost thermal monitor-
per year was observed from the mid 1980s to the mid 1990s
at a depth of IOm in boreholes maintained by Universite Laval
(Allard et al. 1995). This was followed by a warming
trend starting in 1996 (Brown et 2000). al.
SUMMARY
The GTN-P has been established to provide long-term ob
servations of permafrost conditions that will enable us to address
a number of climate change issues in the permafrost
regions. The data collected by this network will provide information
to scientists studying the cryosphere, and also to
stakeholders, politicians and other decision makers. This information
is critical in order to evaluate the impacts of climate
change and to develop adaptation measures to reduce these
impacts.
Results from North America indicate that warming of permafrost
occurred in the latter half of 20th the century in Alaska
and the western Canadian Arctic. Warming of permafrost is
also observed in the Canadian eastem and high Arctic and
mainly occurred in the late 1990s. These trends in permafrost
temperature are consistent with the increases in air temperature
observed since the 1970s in the North American Arctic.
Brown, J., Hinkel, K.M. and Nelson, F.E. 2000. The Circumpolar Active
Layer (CALM) program: research designs and initial results.
Polar Geography 24: 163-258.
Burgess, M.M., Smith, S.L., Brown, J., Romanovsky, V. and Hinkel, K
2000. Global Terrestrial Network for Permafrost (GTNet-P):
permafrost monitoring contributing to global climate observations.
Geological Survey of Canada, Current Research 2000-
E14.
Romanovsky, V. Burgess, M., Smith, S., Yoshikawa, K., and Brown, J.
2002. Permafrost temperature records: indicators of climate
change. EOS, Transactions of the American Geophysical Union
83 (50): 589.
REFERENCES
Allard, M., Wang, B. and Pilon, J.A. 1995. Recent cooling along the
southern shore of Hudson Strait Quebec, Canada, documented
from permafrost temperature measurements, Arctic and Alpine
Research 27: 157-1 66.
154

8th international Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Satellite radar interferometry for detecting and quantifying mountain
permafrost creep
T. Strozzi, U. Wegmiiller, *A. Kaab and R. Frauenfelder
Gamma Remote Sensing, Muri BE, Switzerland
"Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
Remote-sensing techniques represent suitable tools for
integral hazard mapping and monitoring in high mountains,
regions that are typically difficult to access. Spaceborne
Synthetic Aperture Radar (SAR) systems offer the
possibility, through differential SAR interferometry (Dln-
SAR), to map surface displacements at mm to cm resolution
in the line-of-sight direction (Bamler and Hart1 1998).
Recent studies on rock glacier deformation suggest that
the technique is promising also for relatively small objects
in mountainous terrain (Rignot et al. 2002; Kenyi and
Kaufmann, 2003).
We analyzed a series of SAR images between 1993
and 1999 from the European Remote Sensing Satellites
ERS-I and ERS-2 (C-Band, 5.3 GHz, 5.6 cm wavelength)
and the Japanese Earth Resources Satellite JERS-1 (L-
Band, 1.3 GHz, 23.5 cm wavelength). From the large
number of the possible interferometric pairs we considered
interferograms with short baselines acquired during
the snow-free period between early summer and mid-fall.
The topographic reference was determined from two winter
ERS-112 Tandem pairs and subtracted from the interferograms.
The differential interferograms were orthorectified
to the Swiss cartographic system with a pixel spacing
of 10 m by use of a Digital Elevation Model (DEM) derived
from aerial photogrammetry (Kaab 2002). For the most
convincing interferograms the analysis continued with
phase unwrapping permitting the quantitative measurement
of the displacement field of identified rock glaciers.
Phase unwrapping of interferograms in rugged terrain and
for complicated displacement fields is a critical task that
was successful only for small areas. The line-of-sight displacement
was finally transformed in displacement along
the slope direction using the DEM.
Results for the Fletschhorn region in the SimplonlSaas
valley region, Swiss Alps, were analyzed together with the
survey of rock glaciers (Frauenfelder et al. 1998) and displacement
maps from digital photogrammetry of repeated
airborne imagery (Kaab 2002). Two examples are shown
in Figures 1 and 2. For large active rock glaciers such as
the Rothorn (Figure 1) JERS DlnSAR is able to display the
entire flow field with the typical velocity distribution of
highest speeds in the rock glacier center. Below the Jegi
rock glacier (Figure 2) the advantage of DlnSAR for detecting
small movements becomes apparent. Field classification
gave a transition from active over inactive to relict
parts. In the ERS differential interferogram surface deformation
can clearly be seen. It is however an open issue
whether deformations observed from DlnSAR for the inactive
and relict parts of the rock glaciers are due to 'active'
ice deformation or due to 'passive' pushing by the upper
active rock glaciers. Also for the debris slope to the south
of the Jegihorn, which was not classified as rock glacier, it
is not clear if the small movements are due to deformation
of frozen debris.
For rock glacier research and assessment of related
slope instability hazards, our study suggests the following
prior fields of application for DlnSAR: (i) DlnSAR is especially
useful for area-wide detection of rock glacier activity
and order of surface velocity; (ii) the detection of very
small movements by DlnSAR is feasible and will provide
new insights into the behavior of inactive or even relict
rock glaciers; (iii) such results reveal valuable parameters
for further detailed investigations at selected sites by
photogrammetric or field-based methods; (iv) DlnSAR is
able to provide rock glacier speed for a large number of
such landforms and might promote the understanding of
the coupling between landscape, climate and rock glacier
activity.
The major limiting factors of DlnSAR in mountainous
terrain arise from temporal decorrelation and the SAR image
geometry, both leading to incomplete spatial coverage.
Over rock glaciers, where dense vegetation is no
longer present, high coherence is regularly observed during
the snow-free period. Also, large and incoherent displacements
of adjacent scatters may cause decorrelation.
Coherence maps may therefore support the detection of
displacements. The very rugged topography of alpine regions
causes incomplete coverage due to layover and
shadowing. Furthermore, there is a privileged slope direction,
namely that facing away from the SAR look vector,
where the technique is better suited for the monitoring of
displacements.
155

tory afound The Inner ~ o~ho~n from in-situ
nt v ~ l ~ ~ c ~ i ~ from ~ i vsa~~i!ite
e ~
d ~ ma~e is an a ~ r i ~ l i ~ ~ ~ e r y ~
indi~ato~s, such as ~eomorphoio~i~al ~nter~ret~tion~ ~ ~ wa- ~
on of vege~~tion cover^ and ioeation~ of burrows of ~ ib~rna~
, in~€~iv~
dark gr~y), and relict ( l i gray^. ~ ~ Image ~ wid^^ is 7.5
€ui~ition time in~e~al and -65 m perp~n~i~ular ba~eli (c)
ming creep in the direet~on of maximum slope (max~mum 60
the ~e~ihorn from in-situ na~u~al indi€~tors~ ~egend as in Fi~ur~
I. Image width is 1.5 km. (b) ERS
'10.2~ with 105 days a€ui~ition time inte~al and 5 m perpen~i~ular bas~lin~. (c) ~u~~
ffom sate!ii~e radar in~e~e~~metry by assuming creep in the di~~~tion of maximum lo^^ (maxi~um 10 ~ m/y~~r).
Rignot, E., Hallet, B., and Fountain, A. 2002. Rock glacier surface
ACKNOWLEDGEMENTS
motion in Beacon Valley, Antarctica, from synthetic-aperture radar
ERS SAR data courtesy A03-178,O ESA, processing GAMMA. JERS
interferometry. Geophysical Research Letters,
GL013494,29 June 2002.
10.102912001
SAR data courtesy J-2RI-001, 0 NASDA, processing GAMMA. Aerial
imagery taken by Swisstopo and the Swiss Federal Office of Cadastral
Surveys.
REFERENCES
Bamler, R., and Hartl, P. 1998. Synthetic aperture radar interferometry.
Inverse Problems, 14, RI-R54.
Frauenfelder, R., Allgower, B., Haeberli, W., and Hoelzle, M. 1998.
Permafrost investigations with GIS - a case study in the
Fletschhorn area, Wallis, Swiss Alps. Proc. 7th Intern. Conf. Permafrost,
Yellowknife, Canada, Collection Nordicana. 57: 291-295.
Kaab 2002. Monitoring high-mountain terrain deformation from digital
aerial imagery and ASTER data. ISPRS Journal of Photogrammetry
and remote sensing. 57(1-2): 39-52.
Kenyi, L.W. and Kaufmann, V. 2003. Measuring rock glacier surface
deformation using SAR interferometry. 8th Int. Conference on
Permafrost. Zurich.
156

8th International Conference an Permafrost Extended Abstracts on Current Research and Newly Available Information
An attempt to obtain paleoclimatic information from the present permafrost
distribution in Siberia
T. Sueyoshi and *Y. Hamano
Laboratory of Hydraulics, Hydrology and Glaciology, ETH Zurich
*Department of Earth and Planetary Sciences, Univ. of Tokyo, Japan
INTRODUCTION
The thickness of permafrost changes in response to climate
change. Changes in surface temperature are transmitted
through thermal conduction, thus, the response time becomes
long for thick permafrost.
Siberia is a region in which some of the thickest permafrost
is found, sometimes more than several hundreds meters
deep. Such thicknesses of permafrost should reflect the surface
temperature history of the past. In spite of its importance
for the understanding of long-term climatic variability, the terrestrial
paleotemperature history of the continental inland is still
open to argument because most "high-resolution" paleoclimatic
studies are based on marine sediments or ice cores. In
this study, we attempt to obtain paleoclimatic information from
the present permafrost distribution in Siberia by means d
numerical methods.
METHOD
We used a one-dimensional numerical model of permafrost in
which the ground temperature profile is calculated in response
to the variations of surface temperature change from LGM
(Last Glacial Maximum, 21 ka) to the present.
We worked with the one-dimensional thermal conduction
problem, we took latent heat production into account but ne
glected soil water diffusion. Latent heat was considered by
using the enthalpy method, which generally gives good results
when calculating frozenlunfrozen boundary positions.
The thickness of the permafrost is obtained by tracking the
position of this frozen front.
Three study sites for numerical experiments, Tiksi (71"N),
Yakutsk (62'N) and lrkutsk (52"N) were selected in this
study to investigate the paleoclimatic variability in east Siberia.
These sites are located along the Lena River and have
similar geological conditions. The locations, permafrost thickness
and the mean annual air temperature (MAAT) of the
sites are shown in table 1.
Table 1.
Site name Lat. Log thickness MAAT
Tiksi 71.35'N 128.55'E 650m -14'C
Yakutsk 62.05'N 129.45'E 500m- -8KC
lrkutsk 52.16'N 104.21"E 25m- -0.OT
These permafrost thicknesses are based on
and Devyatkin (1974) and Harada (2001).
Nekrasov
Boundary Conditions:
Conditions considered in this simulation are summarized as
follows:
I) Geothermal heafflow value (30 - 60 mW/m2)
2) Local climate condition:
annual mean temperature, seasonal variability
3) Temperature drop during LGM (DT)
a) "normal" paleocliamte scenario:
same as the ice-core reconstruction (DT-10'C)
b) "extreme" paleocliamte scenario:
colder climate (DT -15'C, 20"C, 25°C)
4) Effect of ice sheet existence (only in the Tiksi case)
Top and bottom boundary conditions are necessary to tun
the model. Geothermal heatflow was given for the bottom
boundary conditions and surface temperature was used for
the top.
We assumed the value of the heatflow to be constant for
the calculated period since the study area is on a stable
shield continent. We found this assumption to be reasonable.
Surface ground temperature is substituted by air temperature.
Present climatic data was used when considering the
climatic conditions of the study sites. Paleoclimate history
was obtained from reconstructed temperaturedeviation data
from Greenland ice cores (GRIP, 1993) in order to represent
the outline of climate change. The general pattern of climate
change was assumed to be basically the same for all sites,
while the temperature drop between Holocene and LGM was
treated as a parameter. The blanketing effect of the ice sheet
during the glacial period was also considered.
The effect of the ice sheet on ground temperature was als
taken into account in the model. The ground temperature un-
157

der the ice sheet was calculated from the air temperature d
the surface of ice sheet and assuming a temperature gradient
through the ice sheet. The air temperature on the of top the ice
sheet was calculated using the adiabatic gradient of the atmosphere
(0.6"C/IOOm) and assuming the thermal conductivity
of ice to be 2.2 [Wlklm]. The thickness of the ice sheet
was also treated as a parameter but this condition was considered
only for the northSiberian case. The reconstruction d
ice sheet distribution was based on CLIMAP (1976) and
Pertier (1994).
Sueyoshi, T. et aL: An attempt to obtain paleoclimatic information ...
ness around Tiksi can be reproduced by the "normal" palm
temperature scenario, thereby this "extreme" cooling is
thought to be characteristic for inland Siberia.
1200 , , I I I I I I I ( I . I .
I
I . . . . , . , * , ....
Yakutsk (62'N, 1 JO'E)
RESULTS AND DISCUSSIONS
the
A series of numerical experiments was performed under
various boundary conditions mentioned in the previous section.
We obtained the temporal variations of permafrost thickness
for the last 30k years (from LGM to the present).
0 5000 10000 15000 20000 25000 30000
In this paper, we show the results from the parameter
yearsBP
study of the temperature drop during LGM. The target of these Figure 2. Temporal change of permafrost thickness at Yalutsk.
experiments was to examine the sensitivity of permafrost
thickness to the climate during LGM and to argue to what On the other hand, the distribution of permafrost in the
extent the present thickness of the permafrost has been af- south-marginal area could not be reproduced, even under
fected by past climate history.
"extreme" conditions. Since such thin permafrost responds
Figure1 shows the temporal variation of the permafrost quickly to surface climate change it should not be regarded as
thickness around Tiksi, from 30ka to the present, using sev- a relic of a past cold climate.
eral different values of temperature drop during LGM (DT). In
this plot the X-axis represent the years before present while
the Y-axis represents the thickness of the permafrost. The ar- CONCLUSIONS
row at the Y-axis shows the present thickness of permafrost
at that location. From the figure we can tell that the present The permafrost thickness in the central (Yakutsk) and northem
permafrost thickness in this region roughly agrees with the part (Tiksi) of Siberia cannot be explained consistently by the
"normal" paleoclimate scenario.
same paleoclimatic boundary conditions. The present pemafrost
thickness at Tiksi generally agrees with the value of the
Tiksl(7laN,1 29'E)
1200,....,.., . . . . . . . . . . . . . . . . . . . . . .
calculated thickness, while that at Yakustk requires colder
- 1000
E
Y
U
IC
E
800
0 600
n
c
400
i
............ J. ...... 1 ....... !..".,_"
..I
--.-"'1'"' ! : AT 5 S'C
............................ :. ............!... ... ................................. ......
I
o ~ " " ' " ' ~ ' " " ' " ' " ' ' " ' ' ~
0 5000 10000 15000 20000 25000 30000
yearsBP
Figure 1. Temporal change of permafrost thickness at Tiksi.
Figure 2 shows the case for Yakutsk in the central part of
Siberia.
Comparison with the present thickness suggests that this
area could have experienced enhanced cooling during LGM
which is different to that obtained from the reconstructed ice
core data. If we do not consider this effect the conditions for
this region are too mild, the thickness of permafrost should be
thinner than it is at present. However, the permafrost thick-
conditions. This suggests that the temperature drop during
LGM in the central part of Siberia was larger in the north.
REFERENCES
CLIMAP Project Members. 1976. The Surface of the Ice-Age Earth;
Science 191: 1131-1137.
GRIP Members. 1993. Climate instability during the last interglacial
period recorded in the GRIP ice core; Nature 364: 203-207.
Harada, K. 2001, Study on detection of permafrost structure., DOCtoral
Thesis in Hokkaido Univ.
Nekrasov, LA. and Devyatkin, V.N. 1974. Morphology of permafrost -
underlain areas of the Yana River basin and adjacent regions,
Nauka Press, Novosibirsk (in Russian).
Pertier, W.R. 1994. Ice Age Paleotopography, Science 265: 195-
201.
158

8th International Conference on Permafrost Extended Abstracts on ~~~~~~~
Research and Newly Available lnforrnatlan
Isotopic composition of the massive ground iceof the Bovanenkovo
geotectonic structure (Yamal Peninsula, Russia)
A. M. Tarasov, 1. D. Streletskaya* and N. G. Ukraintseva*
“GROUND” CO. LTD, Moscow, Russia
*Moscow State University, Moscow, Russia
INTRODUCTION
A study of massive ice deposits was carried out within the
framework of an engineering geo-cryological survey in the
area of the Bovanenkovo gas-condensate field. Ten large
ice beds, situated at a distance of 5 to 60 km from each
other, were drilled from top to bottom (Kurfurst et al. 1993,
Tarasov 1990, 1993). The ice and host deposits were
sampled, with an increment of 10-15 cm, to determine
their chemical and isotopic composition.
RESULTS AND DISCUSSION
Morphological features of the massive ground ice
It has been established that the large ice deposits stretch
for 200-800 m and that their thickness varies from 5 to
more than 30 m. They form beds with a horizontal area
that can exceed the vertical by up to a dozen times. Deposits
of this type are widely spread in sections of the Upper
Pleistocene marine plains and lie mainly in the contact
zone between the saline marine clays (Ds=0.3-15%), and
the slightly salinized (Ds=0.02-0.1%) marine-lacustrine
sandy-silt sediments that lie at depths from 1-2 m to more
than 20-30 m (Dubikov 2002). Massive ground ice bodies
are more than 15 m in thickness. In all studied deposits
the bottom of the ice beds is flat, sub-horizontal and lies
beneath the modern sea level (Fig. 1). The contact is consistent
with substrate strike. The ice top is rolling due to
diagenetic processes (melting, thermodenudation and
thermoerosion). Locally, the bedding is concordant with
the covering clay sediments (coarse ice bars). However in
some cases, when the ice is covered by solifluctionlandslide
deposits with cryogenic structure, the contact is
sharply discordant, gently sloping and the ice top is
melted. Ice has been noted with local inclusion of salty
water, cryopegs, voids, cracks and under-ice gases
(Streletskaya and Leibman 2002).
Chemical and isotopic composition of the massive ground
ice
The chemical composition of massive ground ice was
analyzed at study sites 1-88 (central part of the gas field
area). An exposure of massive ground ice was found in a
west slope relic hill in 1987. The bottom of the massive
ground ice body is horizontal and situated at 1-3 m above
sea level. The massive ground ice body is flat-layered, ultra
fresh, and abundant in mineral inclusions. The salt
content in pure ice (melted) is very low and varies from 10
to 80 mgldm3. The salt content in massive ground ice with
mineral inclusions increases up to 160-180 mgldm3.
Water-soluble HC03- is the dominant anion and Na+, Catt
dominated among cations.
The oxygene-18 isotope content (6180) in massive ice
melts is shown in table 1. The background measurements
of 6180 are the same in all ice bodies even though they
can be 30-50 km removed from each other. This proves
that the all massive ground ice in this region has a common
origin. Meteoric water was the source for the formation
of this massive ground ice, marine water did not take
part in this process.
Table 1. Oxygene-18 isotope (8180) in massive ground ice.
Study H a.s.l., Depth of 6180 6180 6180 n*
Site1 m the rn ice, Mean Min Max
Pnint
1-88/2 6.8 1.2-8.7 -18.4 -17.7 -19.7 13
1-8813 10.3 1.8-11.0 -18.0 -17,l -19.0 16
1-8814 12.1 3.7-14.6 -18.4 -18.0 -19.0 20
1-8817 14.4 6.4-15.4 -18.5 -18.1 -19.0 7
2-8811 21.9 6.1-12.0 -18.8 -18.0 -19.7 IO
2-88162 6.4 1.7-7.5 -18.9 -18.6 -19.3 8
2-89122 15.6 3.3-12.8 -18.9 -17.3-20.2 18
3-8911 11.4 3.6-5.3 -19.7 -19.0 -20.6 4
1-9013 7.0 1.0-15.3 -19.0 -17.7 -22.5 16
2-90132 16.4 1.8-10.6 -18.8 -17.6 -20.6 11
2-9011 17.0 1.2-13.2 -18.9 -18.1 -19.3 27
3-9012 20.2 1.9-7.0 -20.6 -16.8 -22.6 12
*n- number of analyzed samples from indicated boreholes
159

Tarasov, AM, et, al.: Isotopic composition of the massive ground ice..,
20 I
A
2a 6 6a 8 12 I 2o
Figure 1. Profiles of geological and ground ice conditions.
A - study sites 1-90, profile 2; B - study sites 2-89, profile 1; C - study sites 2-89, profile 2. 7 - massive ice; 2 - clay, loam; 3 - silty sand;
4 - loamy sand; 5 - hollows in massive ice: 6 - borehole location and number; 7 - lakes.
CONCLUSIONS
Massive ground ice is widespread in the Bovanenkovo
geotectonic structure (Yamal Peninsula).
In all studied deposits the bottom of the ice beds is flat,
sub-horizontal and lie beneath the modern sea level. The
ice top is rolling due to diagenetic processes (melting,
thermodenudation and thermoerosion).
The concentration of oxygen-18 and heavy hydrogen
isotopes in the studied massive ice bodies reveals evidence
of their genetic relation. The obtained results
suggest that the ice had been formed below ground during
the freezing of a water-bearing sandy horizon fed by atmospheric
water precipitation.
Tarasov, A.M. 1993. The isotopic composition of the massive ice of
the Bovanenkovo gas-condensate field. In 'The Isotope in the Hvdrosphere'
Abstr. if the 4th intern. Symp., Pjatigorsk, 18-21 May
1993. Moscow: RAS (in Russian).
ACKNOWLEDGEMENTS
This work was supported by INTAS (grant 01 - 2329) and
RFBR project - 02-05-64263.
REFERENCES
Dubikov, G.I. 2002. Composition and cryogenic structure of permafrost
in West Siberia. Moscow: GEOS Publisher (in Russian).
Kurfurst, P.J., Melnikov, E.S., Tarasov, A.M. & Chervova, E.I. 1993.
Co-operative Russian-Canadian engineering Geology Investigations
of Permafrost on the Yamal Peninsula, Western Siberia. In
Permafrost. Proc. of the 6th intern. Conf. on permafrost, Beijing, 5-
9 July 1993, v.1: 356-361.
Streletskaya, I.D. & Leibman, M.O. 2002. Cryogeochemical interrelation
of massive ground ice, cryopegs, and enclosing deposits of
Central Yamal. Kriosfera Zemli 6(3): 15-24 (in Russian).
Tarasov, A.M. 1990. Experience of applying oxygen-isotope method
to study ground ice at engineering-geological survey. In E.S. Melnikov
(ed.), Methods of engineering-geological survey: 118-133.
Moscow: VSEGINGEO (in Russian).
160

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Mapping of permafrost and the periglacial environments, Cordon del Plata,
Argentina
D. Trornbofto
Affiliation IJnidad de Geocriologia, lnstituto Argentino de Nivologia, Glaciologia y Ciencias Ambientales [IANIGM) - CONICET,
Mendoza - Argentina
Regional mapping of periglacial areas and permafrost in
the Cordillera Frontal, Mendoza, Argentina is unprecedented
in South America - only certain particular sites or
valleys have been mapped. A regional approach was
made, with a focus on analyzing the surrounding limited
periglacial and glaciological environments, and preparation
of an inventory of glacial forms. The resulting map is
intented to form the basis of the geocryological map of
South America.
The study area, Cordon del Plata, lies in the province
of Mendoza, in the Argentine Central Andes between
32”40‘ and 33”24’S and 69”45’ and 69”12’W. Geologically
the mountain ranges belong to the Cordillera Frontal.
The study area ranges from a height of 2000 m a.s.1. to a
maximum height of 6310 m, which corresponds to the
peak of the Cerro El Plata. The minimum height was considered
on the basis of geographical and geological limits
of the area and its independence as a topographical unit.
Three outstanding ranges may be distinguished however:
“La Jaula” in the west, the “Cordon del Plata” in the central
and northern part on the eastern border, and “Santa
Clara” in the southwest. The area comprises ca. 2830
km*.
The MAAT of the meteorological stations of Aguaditas
at a height of 2225 m (1972-1983) and Vallecitos
(1976-1985) at 2500 m at its eastern flank are 7.7”C and
6”C, respectively. The meteorological station Balcon I at
3560 m on the tongue of a rock glacier at Morenas Coloradas
indicates a MAAT of 1.6”C and precipitation of between
500 mm (April 2001 - April 2002), and 630 mm
(1991-1993).
The vegetation ranges from shrubs to the Andean
tundra above 3600 m. Above this zone and on the rock
glaciers vegetation is extremely scarce.
The map has been initiated by scanning and digitizing
topographic maps (scale 1:100.000) using the AutoCad
program Map (R) Release 3. The contour intervals are set
at 100 m. For georeferencing of the obtained data and for
the elaboration of a data base, the program ldrisi for Windows
version 2.010 was applied. As a first step the existing
data were superposed and updated with landsat images.
In a second step the periglacial and glacial data by
Corte and Espizua (1980) obtained from aerial photographs
(scale I: 50.000) between March and May 1963
were translated to the basic map, and corrections applied.
For the periglacial environment studied for mapping the
“facies” (in sense of Corte) of structural debris and inactive
facies were taken from the inventory of glaciers. In the
case of the “thermokarst facies”, it was considered
whether or not they had a direct relation with debris-covered
glaciers. They were included if they belonged to the
environment of rock glaciers (“structural debris facies”) or
to debris-covered glaciers. The 1981 inventory shows that
the debris-covered glaciers preceded the areas or “facies”
with thermokarst, although these are the exceptions. In
those cases where moraine terminals were identified, the
“roots” of rock glacier are found. This allows for the separation
of the two differnet environments.
Permafrost was detected by direct methods, using temperature
profiles in surface drillings (up to a depth of 5 m)
or indirect methods such as geophysical profiling or geomorphological
interpretations by applying aerial photographs
and satellite images. Field visits helped to verify
the interpreted results. For active rock glaciers with southwesterly
orientation the permafrost table was found at
3770 m at a depth of 3 m (Balcon 11).
The periglacial forms and types of permafrost were
mapped according to (Trombotto 2000).
The geomorphological processes according to the altitudinal
levels may be grouped as follows:
1) from 2000 to 3500 m, with periglacial fossil forms,
dynamics of seasonal freezing and semi-arid shaping
of slopes, eventual relict permafrost;
2) semi-arid periglacial level from 3500 to 4500 m, with
different types of permafrost, varied patterned ground,
solifluction forms, glaciers, rock glaciers and mainly
detrital slopes; and
3) snow-covered level, with snow patches, approximately
continuous permafrost, glaciers, rock glaciers and
cryoplanation surfaces (Garleff 1977, Trombotto
1991).
161

With the help of the Andean geomorphologists, involved
and by interpretating the cryogenic or cryotic characteristics
and cryogenic processes or restrictive factors, the
permafrost so far detected by various authors can be
classified and the preliminary typology elaborated (Trombotto
2000). Six'or possibly seven major groups can be
proposed according to their expected ice content, their
water potential, and their genesis (Tab. 1).
The glacier zones have decreased, supposedly in relation
to global change. As a result of the retreat of glaciers
and the thinning of level 3 glacial and snow-covered zone,
the periglacial level has enlarged, and the cryogenic surfaces
occupy new higher zones with moraines and glacial
sediments. However, the variations in the altitudinal periglacial
limit (semi-arid or semi-humid) still have to be defined
more precisely. A reliable catalogue of catastrophic
slope events remains to be compiled. These events have
not been described for Level 2 and are important for Level
1 where human settlements are located. While the area
above 4500 m is likely to have permafrost occurrence
everywhere, the area between 3500 and 4500 m obviously
has fewer possibilities of permafrost and is limited preferentially
to the rock glacier valleys.
Trornbntto: D.: Mapping of permafrost and the periglacial environments
REFERENCES
Corte, A. and Espirlia, 1. 1981. lnventario de Glaciares de la Cuenca
del Rio Mendoza. IANIGLA-CONICET. lmprenta Farras, Mendoza
.
Garleff, K.; 1977. Hohenstufen der argentinischen Anden in Cuyo,
Patagonien und Feuerland. Gottinger Geographische Abhandungen,
Heft 68.
Trombotto, D. 1991, Untersuchungen zum periglazialen Formenschatz
und zu periglazialen Sedimenten in der 'Lagunita del Plata',
Mendoza, Argentinien. Heidelberger Geographische Arbeiten, Heft
90.
Trombotto, D. 2000. Survey of cryogenic processes, periglacial forms
and permafrost conditions in South America. Revista do lnstituto
Geologico 21 : 33-55.
Table 1, Permafrost types and environments (*see Trombotto 2000 for source references).
Dn Andes
Ice Hydric Permafrost Features Restrictors Tropical Desert Central Southern
Content Potential Types andAndes Andes Andes Andes
Processes (17"301-31"S) 131"-35"S) (35"-55"30'S)
+B tfl 1) Creeping Active rock Creeping, Active rock Chile: Rock Argentina: Argentina:
+B +I glaciers, Energy glaciers glaciers active rock active rock
Cold balance, 2175 mmly glaciers as glaciers
Discontinuous Precipitation
indicators in the
sense of Barsch
2) Approxi- Cold Topography Chimborato NW Argentina: Argentina: Argentina:
mate - Energy (6275 m), MAAT -1/-2"C, MAAT-2/4"C, I) 44"s: 2060 rn
Continuous balance, Ecuador:

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
New map of seasonal ground freezing and thawing in Mongolia
(scale1:I
500000)
D. Tumurbaatar, Ya. Jambaljav, D. Undarmaa, M. Enkhtuul and A. Batbold
lnstitufe of Geography, MAS, Ulaanbaatar, Mongolia
INTRODUCTION
In 1971 D. Tumurbaatar compiled a map of seasonal
freezing and thawing of the ground around Ulaanbaatar
city at a scale of 1:25,000. In 1975 S. J. Zabolotnik
created a map of seasonal freezing and thawing of ground
at a scale of 1:2,500,000. In 2001 D. Tumurbaatar
compiled a new map of seasonal freezing and thawing of
ground in Mongolia at a scale of 1 :I 500000. It is based on
40 years of research on permafrost in Mongolia (Devyatkin
1976, Jambaljav 2001 , Kudryavtsev 1978, Tumurbaatar
1973, 1993). For this map we have used data of the
physical and thermophysical properties of 2000 samples
and climate data from 234 meteorological stations located
in the territory of Mongolia.
SEASONAL FREEZING AND THAWING
The following methods were used: (1) field survey and
laboratory test analyses, (2) statistical analyses of field
surveys and laboratory tests, (3) B.A.Kudryavtsev's equation
for the determination of freezing and thawing depths,
(4) The winsurfur software drawing the contours of mean
annual surface temperature (MAST) and mean annual
surface amplitudes of temperature (MASAT).
The map of Quartenary geological deposits of Mongolia,
compiled by E.V.Devyatkin, was taken as a base map.
Depths of seasonal freezing and thawing of ground in
Mongolia vary separately as follows:
1) Depth of seasonal thawing for sandy loam and loam
filling debris of glacial deposits is 2.1 - 3.7 meters, where
MAST is 0 - -2"C, MASAT is 21 - 25°C and ground moisture
is 14 - 29.1%. Where the MAST is 0 - 4"C, the depth
of seasonal freezing and thawing is 1.7 meter in the south
and 3.8 meter in the north of the country.
2) Depth af seasonal thawing in alluvial deposits is 2.4
meter for loam filling debris and 4.7 meter for sand filling
debris, where MAST is -2- O"C, MASAT is 21 - 31°C and
ground moisture is 6 - 22%. Where MAST is 0 - 4"C, MA-
SAT is 21 - 25°C and ground moisture is 6 - 22%, the
depth of seasonal freezing and thawing is 1.9 meter for
sand filling debris in the south and 4.9 meter in the north.
Where MAST is 4 - 7"C, depth of seasonal freezing is I .7
meter in the central parts and 3.5 meter in the south of the
countrv.
3) Depth of seasonal thawing in lacustrine deposits is
2.1 meter in the north, 3.6 meter in the central parts for
clay and loam, where MAST is 0 - -2'C, MASAT is 23 -
25°C and the ground moisture is 14 - 29.1%. Where
MAST is 0 - 4"C, MASAT is 21 - 31°C and ground moisture
is 14 - 29%, the depth of seasonal freezing and
thawing is 1.7 meter in the south and 3.7 meter in the
north. Where MAST is 4 - 9"C, MASAT is 21 - 25"C, the
depth of seasonal freezing is 1.0 meter in the south and
2.8 meter in the central parts of the country.
4) Depth of seasonal thawing in glacial-lacustrine deposits
is 2.3 - 3.7 meter for sandy loam and loam, where
MAST is 0 - -2"C, MASAT is 21 - 25°C and ground moisture
is 14 - 29.1%. Where MAST is 1 - 4"C, the depth of
seasonal freezing is 1.9 meter in the south and 3.6 in the
north of the country.
5) Depth of seasonal thawing in alluvial-lacustrine deposits
is 2.7 - 3.7 meter for sand and sandy loam with debris,
where MAST is 0 - -1 "C, MASAT is 31 "C and ground
moisture is 14 - 25%. Where MAST is 0 - 4"C, MASAT is
23 - 27"C, the depth of seasonal freezing and thawing is
1.9 meter in the south and 3.7 meter in the north. Where
MAST is 4 - 8"C, MASAT is 21 - 25"C, the depth of seasonal
freezing is 1.1 meter in the south and 2.8 meter in
the north of the country.
6) Depth of seasonal freezing in proluvial deposits is
2.3 - 3.2 meter for loam and sandy loam with debris,
where MAST is 1 - 4'C, MASAT is 21 - 27°C and ground
moisture is 7 - 12%. Where MAST is 4 - 9"C, MASAT is
21 - 25"C, the depth of seasonal freezing is 1.3 meter in
the south and 2.6 meter in the central parts of the country.
7) Depth of seasonal freezing in proluvial-lacustrine
deposits is 2.1 - 3.2 meter for loam and sandy loam with
debris, where MAST is 1 - 4"C, MASAT is 21 - 25°C and
ground moisture is 7 - 20.5%. Where MAST is 4 - 9"C, the
depth of seasonal freezing is 1.2 meter in the south and
2.8 meter in the central parts of the country
8) Depth of seasonal freezing in eluvial deposits is 2.7
- 4.1 meter for sand, where MAST is I - 4"C, MASAT is 23
- 31°C and ground moisture is 8%. Where MAST is 4 -
9"C, MASAT is 21 - 25"C, the depth of seasonal freezing
is I .9 - 3.23 meter.
163

9) Depth of seasonal thawing in volcanic deposits is 4
meter and more for basalt, where MAST is -2 - O"C, MA-
SAT is 23 - 25°C. Where MAST is 0 - 5"C, the depth of
seasonal freezing is less than 4.0 meter.
10) Depth of seasonal freezing and thawing in eluvial
deposits is 1.9 - 4.7 meter for debris with loam and sandy
loam fill, where MAST is 0 - 4"C, MASAT is 21 - 29°C and
ground moisture is 7 - 22%. Where MAST is 4 - 9"C, MA-
SAT is 21 - 23"C, the depth of seasonal freezing is 1.1
meter in the south and 3.5 meter in the central parts and
north of the country.
I I) Depth of seasonal thawing in eluvial-solifluction
deposits is 2.7 - 3.4 meter for loam and sandy loam with
debris, where MAST is 0 - -IoC, MASAT is 23 - 25°C and
ground moisture is 7 - 22%. Where MAST is 4 - 9"C, MA-
SAT is 21 - 25"C, the depth of seasonal freezing is 2.3 -
3.4 meter.
12) Depth of seasonal thawing in elluvial-deluvial deposits
is 3.0 - 4.6 meter for debris with sand fill, where
MAST is 0 - -2"C, MASAT is 23 - 25°C and ground moisture
is 7 - 22%. Where MAST is 0 - 4"C, MASAT is 21 -
29"C, the depth of seasonal freezing is 1.9 meter in the
south and 4.6 meter in the central parts. Where MAST is 4
- 9"C, the depth of seasonal freezing is I ,I in the south
and 4.3 meter in the north of the country.
13) Depth of seasonal thawing in delluvial-solifluction
deposits is 3.0 - 4.5 meter for loam and sandy loam with
debris, where MAST is 0 - -2"C, MASAT is 23 - 25°C and
ground moisture is 7 - 20%. Where MAST is 0 - 4"C, MA-
SAT is 23 - 29"C, the depth of seasonal freezing and
thawing is 1.9 meter in the south and 4.4 meter in the
north. Where MAST is 4 - 9"C, MASAT is 21 - 25"C, the
depth of seasonal freezing is 1.1 meter in the south and
3.4 meter in the central parts.
14) Depth of seasonal thawing in bedrock deposits is
more than 4 meter for granites, where MAST is -2 - 0°C.
Where MAST is 4 - 9"C, the depth of seasonal freezing is
less than 4 meters.
Using this new map, we have divided three regions of
seasonal freezing and thawing (Fig. I).
Region of seasonal thawing:
There is continuous and discontinuous permafrost in
this region. The ground thawing begins during the first
decade of May and ends in the last decade of October.
Region of seasonal freezing and thawing:
Isolated permafrost is characteristic of this region
(2°C). The ground freezing begins in the last decade of
October and ends in the last decade of April.
Region of seasonal freezing:
In this region, permafrost does not exist. Ground
freezing begins during the first decade of November and
ends in the last decade of March.
Tumubaatar, D. et. al.: New map of Seasonal ground freezing ...
CONCLUSION
I) On the represented map are shown the MAST, MASAT,
superficial deposits and ground freezing and thawing
depths.
2) Minimum depth of seasonal thawing is 2.1 meter for
clay and maximum depth of seasonal thawing is 4.7 meter
for sand. Mean depth of seasonal thawing is 2.2 meter for
loam and 3.9 meter for sandy loam.
3) Minimum depth of seasonal freezing is 1 .0 meter for
clay and maximum depth of seasonal freezing is 4.9 meter
for sand. Mean depth of seasonal freezing is I .5 meter for
loam and 4.0 meter for sandy loam.
4) The seasonal thawing of the ground begins in the
first decade of May and ends during the first decade of
October. The seasonal freezing of the ground begins in
the last decade of October and finishes in the last decade
of April in the north and central parts of the country. This
freezing begins in the first decade of November and ends
during the last decade of March in the south of the country.
Laks,shp
Figure 1. Schematic map of permafrost regions
REFERENCES
Devyatkin, E.V. 1976. Map of Quartenary geological deposits of Mongolia.
Jambaljav, Ya. 2001. Regression equation of maximum and minimum
monthly mean ground temperature curves. In proceedings of the
International Symposium on mountain and arid land permafrost;
33.
Kudryavtsev, V.A. 1978. General geocryology (In Russian).
Tumurbaatar, D. 1993. Seasonal freezing and thawing grounds of
Mongolia. In proceeding of International Conference on permafrost.
Bejing, China, V01.2; 1242-1246.
Tumurbaatar, D. 1973. Seasonal freezing and thawing grounds of
Mongolia. (In Mongolian).
164

8th International Conference on Permafrost Extended Abstracts on Currmt Research and Newly Available Information
Decadal changes in permafrost, land form and land cover near Barrow,
Alaska
C .E. Tweedie, R. D Hollister and P. J. Webber
Department of Plant Biology, Michigan State University, East Lansing, USA
INTRODUCTION
A significant challenge in global change science is assessing
the impact and feedbacks of climate change and
land-uselland-cover change on ecosystem function,
namely carbon exchange across the land-atmosphere
boundary. In the Arctic, terrestrial carbon balance is regulated
by complex spatial and temporal interactions between
land cover, climate, land form, hydrologic balance,
permafrost, and stochastic events such as fire, herbivory
and human disturbance. The importance of arctic regions
to global energy balance and biogeochemical and hydrological
cycling is well recognized, as is the documented
evidence, propensity and sensitivity for these to change
(sometimes non-linearly) in response to climate change
(Hinzman et al. submitted). The IPCC (2001) forecast Arctic
regions as continuing to undergo among the fastest
rates of climatic change on Earth over the next century.
Clearly, the need for understanding the interaction between
permafrost, land form, hydrologic balance, land
cover and the response of ecosystem function in Arctic
regions to climate change and following surface disturbance
has never been greater.
Most investigations of the functional response of Arctic
ecosystems to climate change have been derived from
manipulative experiments, modelling, and paleoenvironmental
investigations with few studies illustrating a
thorough and multidisciplinary examination of the interaction
between ecosystem properties and processes. Limited
by the availability of adequate and long-term datasets,
few forecasts of ecosystem function in Arctic regions have
been validated by observed changes. In the absence of
long-term and integrated multidisciplinary datasets, reoccupation
and retrospective analysis of research sites
established several decades ago offer the best means by
which decadal change in ecosystem function can be assessed
or inferred.
This report details careful re-sampling of vegetation
plots established at Barrow, Alaska during the Tundra Biome
Project of the International Biological Program (IBP
1969-74). Results are discussed in light of other studies in
the Barrow region documenting anthropogenic disturbance
(Brown et al. 1980 and references therein) and
changes in biogeochemical cycling (Oechel et al. 1995,
2000).
METHODS
Plots varied in size from a few square meter plots nested
within single land cover units to a 4.3 hectare grid spanning
the majority of land cover types in the Barrow region.
Following relocation of plot markers between 1999 and
2002, resampling was performed using the original sampling
methods. Comparison of seasonal active layer progression,
microtopography, vegetation composition and
abundance and high resolution land cover maps several
decades apart has allowed the direction and magnitude of
change in permafrost, land form, hydrologic balance and
land cover to be assessed.
RESULTS
Seasonal active layer progression and maximum end of
season active layer depths recorded between 2000 and
2002 were on average 2 to 10 cm greater than those recorded
across all plots in the early 1970s. Comparison of
the range in relative elevation of the ground surface
measured at marked grid points between 1973 and 2000
suggest that there has been an increase of up to 10-15cm
between polygon trough bases and adjacent rims. At the
4.3 ha landscape level, however, there has been a decrease
of 30 cm in the range of relative elevation.
Change in vegetation was greatest in plots established
in formerly disturbed or early successional communities
situated on alluvial substrates. In undisturbed communities
on non-alluvial substrates the greatest change were observed
in wet and moist tundra. Dry tundra displayed little
change over time. At the landscape scale, wet and moist
tundra classes show a reduction in aerial coverage
whereas dry tundra classes show an expansion in area.
Within the 4.3 ha plot, spatial heterogeneity in land cover
increased due to a decrease in patch size and increase in
patch number.
165

DISCUSSION
Tweedie, C.E. et 31.: Decadal changes in permafrost, Barrow, Alaska.
REFERENCES
Differences between measurements made soon after plot
establishment and almost three decades later are dramatic.
Elevations lack registration relative to points of
known height above sea level in initial measurements,
which creates difficulty in assessing whether subsidence
or soil development (accumulation of organic material) can
be attributed to the decrease in the range of relative elevations
recorded at the landscape level. Nonetheless, the
minimal accumulation of organic matter around marker
stakes in the most productive wet tundra land cover
classes suggest that the observed change is due to subsidence
of polygonized tundra in the gradient studied. The
increase in the range of relative elevation between polygon
troughs and adjacent rims suggests thermokarsting of
ice wedges underlying polygon troughs. Change in vegetation
and land cover also supports this hypothesis.
Similar patterns of thermokarst development have
been observed in nearby coastal zones where erosion of
polygonized tundra is occurring. These elevation changes
are potentially significant considering that most of the
study area and nearby settlement of Barrow are ca. 3
meters above sea level, indicating up to a 40% decrease
in elevation relative to sea level and increased susceptibility
to storm surges and flooding.
Within the same study area, Oechel et al. (1995) have
documented a change from tundra being a net sink for
C02 to the atmosphere in the early 1970s to being a net
source by the early 199Os, implying a positive feedback
response to climate change. The increased active layer
thickness and change in land cover documented in this
study and the well-studied relationship of CO2 sink activity
declining in response to increasing water deficit provides
evidence of mechanisms driving patterns documented by
Oechel et al. (1995). Although rates of water deficits appear
to have increased during the snow-free period in the
region over the past few decades (Oechel et at. 1995
2000) and cannot be discounted, we suggest that the
above results can be most strongly attributed to a shift in
hydrologic balance caused by anthropogenic disturbance
and altered land use patterns. These have been dramatic
near Barrow since the mid 1940’s.
Brown, J., Miller, P.C., Tieszen, L.L. and Bunnell, F.L. (eds). 1980. An
Arctic Ecosystem: The Coastal Tundra at Barrow, Alaska.
Stroudsburg: Dowden, Hutchinson & Ross Inc.
Hinzman, L. et al. (29 other authors submitted 2003). Evidence and
implications of recent climate change in terrestrial regions of the
Arctic. Climatic Change.
IPCC. 2001. Climate Change 2001. The Scientific Basis. Contribution
of Working Group I to the Third Assessment Report of the Intergovernmental
Panel on Climate Change. In Houghton, J.T., Ding,
Y., Griggs, D.J., Noguer, M., van der Linden, P.J., Dai, X.,
Maskell, K. and Johnson, C.A. (eds.). Cambridge: Cambridge
University Press.
Oechel, W.C., Hastings, S.J., Vourlitis, G.L., Jenkins, M.A., Riechers
G.H. and Grulke, N.E. 1993. Recent change of Arctic tundra ecosystems
from a net carbon dioxide sink to a source. Nature 361:
520-523.
Oechel, W.C., Vourlitis, G.L., Hastings, S.J., Zulueta, R.C., Hinzman,
L. and Kane, D. 2000. Acclimation of ecosystem C02 exchange in
the Alaskan Arctic in response to climate warming. Nature 406:
978-981.
CONCLUSION
This study has illustrated significant changes in ecosystem
properties over the last thirty years near Barrow, Alaska
and has linked these changes to documented changes in
ecosystem function in the region. Ongoing research is investigating
the relative importance of climate change, successional
change (thaw lake cycle) and anthropogenic
disturbance on ecosystem function in the Barrow area. It
is proposed that the former IBP Tundra Biome Project
study area and the adjacent Barrow Environmental Observatory
(BEO) become key sites in the developing Circum-Arctic
Environmental Observatories Network
(CEON).
166

8th International Conference on Permafrost Extended Abstracts an Current Research and Newly Available Information
Contemporary geotechnologies providing safe operation of railway
embankments in-permafrost conditions
V. Ulitsky, V. Paramonov, S. Kudryavtsev ,K. Shashkin and *M. Lisyuk
St. Petersburg State University of Highways: Moskovsky Pr. 8, Saint Petersburg, 190031, Russia
*Georeconstrucfion Engineering Co ,Izmaylovsky Pr. 4, Saint Petersburg, 198005, Russia
Frost heave and thaw related deformation of ground on
Russian railways in permafrost conditions and under the
influence of deep seasonal frost penetration attract plenty
of attention as of the earliest moments of construction and
operation of railway subgrades. In order to eradicate actual
deformations and stabilise subgrade soils in heave
sensitive areas foam polystyrene panels are currently employed
(according to the following pattern: effectiveness
equals costs plus technological effectiveness plus durability).
This method of embankment reconstruction is currently
construed to be of prime efficacy in potentially severe
conditions. Design of such protective measures
requires assessment both of thermophysical characteristics
and of stressed-strained ground conditions during
freeze-thaw process.
To solve problems of the above-designated type we
developed a numerical modelling methodology of such
problems incorporated into FEM models software created
by geotechnical engineers of St. Petersburg (Ulitsky
2000). It allows solution of freeze-thaw problems by
means of a thermal conductivity equation with consideration
for phase transformations in below zero temperatures
range for unsteady heat conditions in 3-D ground medium.
In frozen rock phased water (or liquid solution) content
varies within a whole range of negative temperatures. This
variation occurs under the principle of the dynamic equilibrium
state first described by N.A. Tsytovich (Tsytovich
1979). In accordance to this principle the amount of unfrozen
water for a certain type of non-saline ground is established
based on sub-zero temperature values. As was
proved through experiments, a humidity increase in a
ground sample considerably higher than the humidity of
top plasticity border in frozen state leads to unfrozen water
content increase on account of infinitesimal water films
found on ice crystals formed during freezing.
Our FEM-models software features a function of liquid
water content in the ground to assess latent heat of
phased transitions being absorbed or released by ground
depending on ground water phase fluctuations in belowzero
temperatures. Migratory flow directed towards the
freeze front at varying distance between the latter and the
ground water level (GWL) is established based on ex-
perimental data processing. Ground water level location
and the average trend thereof in wintertime are assumed
based on reference hydrometric materials specific for the
given area (GWL - maximum for a monitoring period;
trend - minimum for same period).
The stress-strain state is assessed based on temperature
fields allowing seasonally dependent definition of
deformations and forces in subgrade structures and facilitating
the anticipation of procedures targeted at the
elimination of heave related adversary effects on the superficial
elements of railway tracks. Currently it is possible
to establish ground freeze-thaw development as well as to
define deformation values of heave related uplift and thaw
related settlement of any structures encountered within
the ground bulk.
We performed numerical modelling of freeze-heavethaw
sequence on railway embankments of different
heights by generalised cross-section profiles established
over a period of long-term operation of railways in permafrost
soils of Russian Transbaikalia and the Far East.
For the studied Transbaikalia section of the Russian
railways (station Mogocha, average annual subgrade
temperature being in the order of 0-1 .Oe) foam polystyrene
insulation allowed considerable temperature decrease on
top of the embankment in winter time, however, with no
guaranteed heave elimination. Embankment calculations
for different polystyrene application techniques showed
that the most effective application was that of 0.12 m and
0.18 m on the top section of subgrade and the bank respectively.
Underneath the insulation panels in the embankment
body freezing temperature changed from -2.9 'C to -6.9'C
(Fig. I), thus remaining within the heave temperature
range. In warmer periods the applied insulation precludes
thaw on top of subgrade thus ruling out embankment deformation
in spring-summer time and positively conditioning
stressed-strained subgrade structure (Fig. 2).
To ensure safe operation of embankments in permafrost
conditions with heaving ground affected by seasonal
frost penetration it would be expedient to render geotechnical
forecast by means of numerical modelling. Based on
such forecast one could advise effective geotechnologies
167

to apply in subgrade ground stabilisation both for seasonal
frost penetration and thaw.
Ulitsky, V. el al.: Contemporary gc otechnolagies providing safe operation ...
REFERENCES
Ulitsky, V.M., Shashkin, A.G., Shashkin, K.G. and Paramonov V.N.
2000. FEM models software for modelling and solution of problems
in construction and reconstruction by FE
method.//Reconstruction of cities and geotechnical engineering/
Saint Petersburg. 2 (in Russian).
Ulitsky, V.M., Paramonov, V.N., Sakharov, 1.1.) Kudryavtsev, SA.. and
Bakenov H.Z. 2002. Calculation of frozen foundations footings on
saline heave sensitive soils with thermal insulation. Proceedings
of the international coastal geotechnical engineering in practice.
21-23 May 2002. Atyray, Kazakhstan:l45-147.
Tsytovich, N.A. and Chronic, J.A. 1979. interrelationship of the principal
physicomechanical and thermophysical properties of coarse
grained frozen soil. Bochum, 1978. - Eng. Geol. 13: 163-167.
Figure 2. Epures of temperature distribution in 2 m high embankment
and subsoil thereof with applied insulation in permafrost climate conditions
of station Mogocha in Russian Transbaikalia for the month of
September: 1 - extruded foam polystyrene; 2 - thawed ground;
3 -frozen ground.
168

li,,
8th International Conference un Permafrost Extended Abstracts on Current Research and Newly Available Information
A new approach for interpreting active-layer observations in the context of
climate change: a West Siberian example
A. A. Vasiliev, N.G. Moskalenko and "J. Brown
Earth Cryosphere Institute, Russian Academy of Sciences, Moscow, 119991 Russia
* lnfernational Permafrost Association, Woods Hole, MA 02543 USA
The Circumpolar Active Layer Monitoring (CALM) program
provides spatial and temporal observations of seasonal
thaw depths at more than 100 sites located in different
bioclimatic and landscape conditions of the polar regions
(Brown et al. 2000). Many of the sites consist of 100m to
1000 m grids with considerable terrain variation within
each grid. Thus the interpretation of these data is complicated
because of the diversity of bioclimatic and landscape
conditions found within and between monitoring
sites. Different landscapes may demonstrate different responses
to past, current and future climate changes.
Our analysis, using data from two West Siberian sites, is
aimed at developing a new approach to the interpretation
of active layer monitoring data. This current report was
stimulated in discussions during the CALM synthesis
workshop held in November 2002 at the University of
Delaware and at recent annual conferences on Earth
Cryosphere held in Pushchino, Russia.
There are three CALM grids in West Siberia: Marre-
Sale (R3,69'43' N, 66'52' E); Vaskiny Dachi (R5, 70"17'
N, 6834' E); and Nadym (Rl, 65'20' N, 72'55' E). The
Marre-Sale and Vaskiny Dachi sites are located in typical
tundra; the Nadym site is located in northern taiga. For
this report we use data from the Marre-Sale and Nadym
sites only, because there are no long-term weather records
for the Vaskiny Dachi site. In addition, we use data
from previously established auxiliary sites located in the
vicinities of CALM grids and having much longer observational
records. Active layer depths are considered within
the dominant landscapes of the tundra zone (R3) and the
taiga zone (RI). For Marre-Sale these include peatland,
mire, and five sites within the 1000m CALM grid: dry tundra,
ravine, mire, wet tundra, and sand field as described
by Melnikov et al. (1983). V. Dubrovin and N. Moskalenko
established several of these sites prior to the CALM program
(Vasiliev et al. 1998; Moskalenko et al. 2001).
Figure I illustrates the dependence of the active layer
depth in these landscapes on the sum of positive air temperatures
(DDT, degree-days T> O'C) at the Marre-Sale
and Nadym sites. As a first approximation, the linear dependence
is used. The slope of the straight line approximation
is different for each landscape. We propose that
this slope characterizes the active layer response of a
particular landscape to climate variation. The maximum
inclination is typical of mires (sites 2 and 5); whereas the
minimum inclination is typical of flat peatlands (site I). All
other landscapes are characterized by intermediate values
of active layer depth versus DDT. As the absolute values
of active layer depth are different, it is reasonable to use
relative values for comparing responses of different landscapes
to climate change with the use of a unified scale.
zoo0
DDT,, Marrc-Salc DDT,, Nadym DDT
O "-I,; ~
Figure 1. Correlation between the active layer depth (ALD), and DDT
for the Marre-Sale and Nadym sites (1- peatland, 2- mire, 3-dry tundra,
4- ravine, 5- mire, 6- wet tundra, 7- sand field. Sites 3- 7 are Iocated
on the CALM grid).
The mean value of DDT calculated from long-term air
temperature records can be used as a climatic standard
for a particular site (DDTst). For the Marre-Sale weather
station, DDT,t equals 566C degree days (1914-2002) and
for the Nadym weather station it reaches 1352C (1965-
2002). In Figure I the vertical lines for DDTst represent
this long-term average. A standard active layer depth
(ALDst) for a particular landscape is represented at the in-
169

tersections of this perpendicular with the corresponding
approximation curves. These values are shown along the
ordinate axis (ALDlst to ALD7$t) for Marre-Sale on the left,
for Nadym on the right. These are very significant values
and can be considered theoretical values of mean active
layer depths calculated for particular landscapes. It should
be pointed out that these values can be obtained on the
basis of relatively short-term observation records (5-10
years). These are the mean active layer depth (ALDst and
average active layer depth for observation time (ALL,) in
Table I.
Table 1. Mean active layer depth (ALDst) and average active layer
depth (ALDa,) for the observation period. Duration of observations enclosed
in brackets.
Vasiliev, A. A. et al.: Active-layer observations, climate change, West Siberian..
(3) The diagram showing the dependence of the active-
layer depth on the annual sum of positive air temperatures
in relative units can be used for assessing current
tendencies of climate variation or changes.
If the values calculated on the basis of actual observational
data are generally higher than l.Oj a tendency for
climate warming is observed; the reverse is true when calculated
values of the relative active layer depth are lower
than 1.0. This diagram can also be used for estimating
changes in the active layer depth under specific scenarios
of climate change.
*Landscapes are given in Figure 1 captions
On the basis of these data, we can recalculate all
available data on the ALD and DDT into relative units using
corresponding standard values, and plot another diagram
showing the dependence of the relative active layer
depth on the relative sum of positive summer temperatures
(Fig. 2). Calculations have been made for contrasting
landscapes (mires and peatlands). The resulting
curves for the other landscapes should occupy intermediate
positions. This diagram makes it possible to assess
current tendencies in the active layer development (Fig.
2). If most of the observational data tends to be concentrated
in the right part of the diagram, then this indicates
climate warming and an increase in the active layer depth.
By contrast, the concentration of data in the left part of the
diagram attests to a cooling trend with a corresponding
decrease in the active layer depth.
Similar calculations have been made for the Nadym
monitoring site (northern taiga zone). It is interesting that
the resulting diagrams for mires and peatlands have the
same shape. The response of these landscapes to the
degree of summer warming (DDT values) does not depend
on the climatic zone (Melnikov at al., this volume).
The following conclusions can be stated from this
analysis:
(I) The suggested method makes it possible to calculate
mean maximum depths of the active layer for longterm
periods on the basis of relatively short-term observational
records.
(2) The response (as manifested by the active layer
depth) of seven dominant types of landscapes of the typical
tundra zone and two dominant types of landscapes of
the northern taiga zone to climate changes is evaluated
quantitatively. The active layer depth in mires is very sensitive
to climate changes; the least sensitive is typical of
peatlands.
" -
Coolilg
Warmk g
.. . . ".
0,s 0,7 09 I,l 1,3 1,s
DDT rel.
Max.T, mire - Max NT, mire
Mi. T, peatland - - Min NT, pealand
Figure 2. Diagram of the active layer response to warming and cooling
in tundra (T) and northern taiga (NT) of West Siberia.
REFERENCES
Brown, J., Hinkel K.M., and Nelson, F.E. 2001. The Circum-polar active
layer monitoring (CALM) program: research designs and initial
results. Polar Geography 24(3):165-258.
Melnikov, E.S. (ed.) 1983. Landscapes of a permafrost zone in the
West Siberian gas province. Novosibirsk: Nauka.
Melnikov, E.S., Leibman, M.O., Moskalenko, N.G., and Vasiliev A.A.
2003. Comparative analysis of active layer monitoring at the
CALM spatial variability at European Russian Circumpolar Active
Layer (CALM) sites (this volume).
Vasiliev, A.A., Korostelev Yu.V., Moskalenko N.G., and Dubrovin V.A.
1998. Active layer monitoring in West Siberia under the CALM
program (database). Earth Cryosphere ll(3): 87-90.
Moskalenko, N.G., Korostelev, Yu.V., and Chervova, E.I. 2001. Active
layer monitoring in northern taiga of West Siberia. Earth
Cryosphere V(1): 71-79.
170

Influence of seasonal active-layer thickness on broad-scale vegetation
dynamics in the permafrost zone
S. Venevsky
Obukhov Institute for Atmospheric Physics, Moscow, Russia and Potsdam Institute for Climate Impact Research, Potsdam,
Germany
SPECIFIC BIOLOGICAL AND PHYSICAL
PROCESSESIN THE PERMAFROSTZONE
Change of active-layer thickness
Continuous and discontinuous permafrost is the factor most
unique for the tundra and taiga environment. Dynamics of the
permafrost table, or active-layer thickness depend on the local
surface energy balance, reflecting above all air temperature
as modified by soil thermophysical properties, vegetation,
snow and solar radiation. In particular, the presence of a
thick forest floor and organic peat layer significantly lowers
soil temperatureand active-layer thickness, as low thermal
conductivity of the organic matter effectively insulates the
mineral soil. A snow cover in autumn limits frost penetration,
while in spring it functions as an insulating blanket to delay
thawing of soil active-layer. Active-layer thickness and low
soil temperature provides a strong feedback on vegetation at
high latitudes. Low soil temperatures influence forest prcduG
tivity by restricting available nutrients due to low dewmposition
rates and by reducing water uptake by trees due to an
increase in root resistance and the viscosity of water. Trees
in spring and the first half of summer can also be subject b
severe water stress by high evaporation demands when their
roots are still frozen and cannot provide therequiredwater
uptake.
SIMULATION OF VEGETATION DYNAMICS IN THE
PERAMFROST ZONE
Ecological objective of the study
Annual and seasonal variation of active-layer thickness affects
local soil moisture status and, thus regulates the water
available to plants and fuel moisture in fire-prone vegetation
communities, especially in spring. One should investigate the
role played by seasonal and annual variation of active-layer
thickness in regional fire regimes, vegetation composition and
carbon stocks on a broad scale in the permafrost zone.
Basic vegetation model
The Lund-Potsdam-Jena Dynamic General Vegetation Model
(LPJ-DGVM) (Sitch et al. 2003) combines mechanistic
treatments of terrestrial vegetation dynamics, carbon and
water cycling in a modular framework. Key features include
fast feedbacks between photosynthesis and water balance
routines and representations of slower ecosystem-level processes
including vegetation resource competition, growth, establishment
and mortality, soil and litter carbon turnover. To
link the fire regime and its effects on vegetation dynamics, a
general fire model was incorporated into the LPJ-DGVM.
This model estimates fire conditions based on soil moisture.
Forest fires
After climate and soil, fire is the most important natural factor
influencing nutrient cycling, chemical and physical processes
Principal method of modification of the Vegetation model for
the permafrost zone
In order to meet the objectives of the study, the LPJ-DGVM
in soil and forest dynamics (regeneration, growth, species was revised and considerably modified for the permafrost
composition and mortality) in a boreal part of the permafrost zone by including the active-layer thickness as one of the
zone. Depending on flame length and severity of fire, vegeta- key model variables (Venevsky 2001). Climate variables
tion and topography characteristics, and presence or absence (daily temperature, precipitation and incoming radiation) define
of permafrost, the influence of fire on forest ecosystem is active-layer thickness, soil moisture status and vegetation
highly variable. Frequent fires can radically change the local status. A daily value of the active layer thickness affects
biogeochemical status, redistribute and mobilize nutrients and water uptake by roots, runoff and percolation of water in the
remove the forest floor layer. Therefore, fire in the boreal zone soil and, therefore, vegetation production and soil moisture
can be considered as a natural rejuvenating force, which content. The daily soil moisture content determines the prob
rapidly returns climax forests to the initial stage of their devel- ability of fire, vegetation production status and determines
opment cycle.
thermophysical properties of soil (and as a result, a current
active-layer thickness). Photosynthesis and water uptake by
171

plants are joint processes which significantly influence daily
soil moisture content. Vegetation can be removed by fire in
the model and affects fire by an amount of accumulated litter
(fuel), The dynamics of the permafrost table, or active-layer
thickness, depends on the local surface energy balance, reflecting
first of all air temperature as modified by soil therm@
physical properties, vegetation, snow and solar radiation
(Kudryavtsev et al. 1974). Due to variations in snow cover
or vegetation type and corresponding changes in heat transport
and radiative energy exchange, surface temperatures
may differfromtheairtemperature by several degrees. A
simple periodic trigonometric function with the maximum annual
thaw depth as an amplitude was applied to represent
daily depth profile of seasonal thawing and freezing, which
starts from above. Three regions with measured seasonal
thawing depth were taken for validation of the active layer
model, two in Siberia (Central Jakutia) and one in Alaska
(Kuparuk river basin).
RESULTS OF SIMULATION WITH HISTORICAL
CLIMATE DATA
Set of experiments
To assess the influence of permafrostandfireregimes for
vegetation structure and carbon poolslexchange, the modified
version of the LPJ-DGVM was run in two modes, with and
without active-layer processes. For the global historical
simulation within the LPJ-DGVM, the CRU05 1901-1995
0.5"x0.5" monthly climate data, provided by the Climate
Research Unit, University of East Anglia, UK, was used.
Venevsky. S.: Active-layer thickness, vegetation, dynamics ...
Comparison with observations
Aboveground vegetation biomass differs rather significantly in
the runs with and without active-layer processes. The vegetation
carbon content in the Asian part of Russia changed from
7.5-12.5 kgClm2 to 2.5-10 kgClm2 and generally decreased
1.5-2.5 times, when the processes seen in the permafrost
zone were included in the LPJ-DGVM. The fire return intervals
in western and eastem Siberia decreased two to four
times, especially in the continental areas, in comparison with
the unmodified model version. The average growing stock of
Larch stands by the nine administrative units of Siberia,
simulated by the two LPJ-DGVM versions for the year 1993,
was compared with the data of Russian Forest State The
area of the administrative units varies from 0.4 to 3 IO6km2.
The mean square relative deviation of simulated against ob
served growing stock was 26% for the modified LPJ-DGVM
version, compared with 70% accuracy obtained with the
original version. The growing stock for the largest northem
administrative units, Magadanskaia oblast', Sakha (Jakutia)
and Evenkiiskaia A0 decreased 1.5 to 2 times after the
model modification and became close to reality.
CONCLUSION
The significant improvement of results for the vegetation carbon
in the permafrost zone is mainly influenced by changes
in the hydrological regime with associated transitions in the
fire regimes. Therefore, interaction of fire regimes and
thawlfreeze processes are very important for vegetation
structures in the permafrost zone.
REFERENCES
Venevsky, S. 2001. Broad-scale vegetation dynamics in northeastern
Eurasia - observations and simulations, Universitat fur
Bodenkultur, Vienna, 150 pp.
Sitch, S., Smith, B., Colin Prentice I. and LPJ consortium. 2003.
Evaluation of ecosystem dynamics, plant geography and terrestrial
carbon cycling in the LPJ Dynamic Global Vegetation
Model, Global Change Biology 9: 161-185.
Kudryavtsev, V.A., Garagulya, L.S., Kondrat'yeva, K.A. and Melamed,
V.G. 1974. Fundamentals of frost forecasting in geological engineering
investigations. Nauka, Moscow, 431 pp.
172

Permafrost Monitoring in Switzerland (PERMOS) in the pilot stage
D.S. Vonder Muhll, *R. Delaloye, ** W. Haeberli, **M. Hoelzle and ***E Krummenacher
Delegate for Permafrost of the Glaciological Commission; Universities of Basel and Zurich, Switzerland
*Institute of Geography, University of Fribourg, Switzerland
**Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
***Geotest AG, Davos, Switzerland
INTRODUCTION
The variation of the glacier tongues has been observed
systematically in Switzerland since the 1850s. The Swiss
Academy of Sciences (SAS) and the Swiss Alpine Club
(SAC) were the driving institutions behind ensuring and
maintaining the measurements. Permafrost is a thermal
phenomenon of the subsurface, and with the snow it complements
the cyrosphere. In contrast to glaciers and snow,
systematic scientific investigations of Alpine permafrost
started some 30 years ago.
The drilling through the Murtel-Corvatsch rock glacier
in 1987 triggered a number of follow-up and parallel activities
in permafrost research at several Swiss institutes.
Concepts for a Swiss monitoring network on permafrost
were developed (Haeberli et al. 1993, Glaziologische
Kommission 2000) parallel to efforts initiated by the International
Permafrost Association (IPA, Frozen Ground
1995; Brown 1997). While the IPA monitoring strategy was
developed mainly for high-latitude plain permafrost areas,
PERMOS is adapted to the low-latitude mountain perrnafrost.
development can be analyzed and monitoring parameters
adapted if necessary
During the pilot phase, PERMOS comprises IOboreholes
(which all were drilled within scientific projects with
mainly non-monitoring aims) and 10 BTS areas (some of
which have been observed for several years already) plus
about two flights per year for aerial photographs. The distribution
of the sites is somewhat random (Fig. 1). Spatial
gaps will be filled after the pilot phase.
OBJECTIVES
The main goal is to document status and long-term variations
of Alpine permafrost. PERmafrost Monitoring in
Switzerland (PERMOS) is based on three elements:
1. recording thermal changes in permafrost, in particular
permafrost temperatures in boreholes,
2. determining the lateral distribution pattern near the
lower permafrost boundary using the BTS (Bottom
Temperature of the Snow cover in March) - method in
combination with single-channel temperature-loggers,
and
3. taking aerial photos to build an archive for photogrammetrical
determination of geomorphologic surface
changes due to changes in permafrost characteristics.
As a consequence, the processes involved can be investigated
and modelled to improve the understanding
and knowledge of Alpine permafrost. In addition, long-term
Figure 1. Distribution of the PERMOS sites (drillings and BTS areas)
during the pilot phase in the contours of Switzerland.
MANAGEMENT SETUP
The network was initiated by the Swiss Adhering Body of
the IPA, and the concept (Glaziologische Kommission
2000) approved by the Glaciological Commission of the
Swiss Academy of Sciences (SAS). During the pilot
phase, PERMOS is financed by the SAS, the Swiss Forest
Agency and the Federal Office for Water and Geology.
Measurements are done by eight institutions (Academia
Engadina Samedan, ETH Zurich, the Universities of
Basel, Bern, Fribourg, Lausanne and Zurich, SFISAR Davos).
Annual reports are published by the Glaciological
Commission of the SAS.
173

INITIAL RESULTS
Permafrost temperatures have been recorded at Murtel-
Corvatsch since 1987 (Fig. 2). Processes and causes for
variations are discussed by Vonder Muhll (2001). Similar
temperature series are recorded at all PERMOS drill sites.
Permafrost temperature at IOm depth range is between
-2°C and 0°C. Thickness of the active layer is determined
where a data logger records temperatures.
The BTS-area of Alpage de Mille (2220 to 2460 m
a.s.1.) has been observed systematically since 1996.
Every year at around the same date, BTS values are
measured at more that 60 points to investigate variations
from one year to another. While the pattern is about the
same every year, BTS values vary up to 3°C from one
year to the next (Vonder Muhll et al. 2001). The BTS
working group of PERMOS is analyzing the set up of the
BTS areas and adapting the method for the second pilot
phase.
Due to weather conditions, no aerial photos were taken
in 2001. In 2002, photos were taken in the Upper Engadin
region (Corvatsch-Murtel).
OUTLOOK
by Swiss Federal Departments, together with the Swiss
Glacier Network, in about three years. It is important that
the measurements are taken, interpreted and coordinated
by the scientific community of the Swiss Adhering Body of
I PA.
REFERENCES
Brown J. 1997. Disturbance and recovery of permafrost terrain. In:
Disturbance and recovery in Arctic lands (Ed. R. M. M. Crawford).
Kluwer Academic Publishers: 167-178.
Glaziologische Kommission 2000. Aufbau eines Permafrost-
Beobachtungsnetzes in der Schweiz - PERMOS (Konzept und
Anhang). 5 plus appendix.
http://www.unibas.ch/vr-forschung1PERMOS
Haeberli W., Hoelzle M., Keller F., Schmid W., Vonder Muhll D. and
Wagner S. 1993. Monitoring the long-term evolution of mountain
permafrost in the Swiss Alps. Sixth International Conference on
Permafrost, Beijing, Proceedings 1: 214-219.
Vonder Muhll D. 2001. Thermal variations of mountain permafrost: An
example of measurements since 1987 in the Swiss Alps. In Visconti
et al (eds.): Global Change in Protected Areas.(Kluwer Academic
Publisher): 83-95.
Vonder Muhll D., Delaloye R., Haeberli W., Holzle M and Krummenacher
B. with contributions of 13 others, 2001: Permafrost Monitoring
Switzerland PERMOS, 1, Jahresbericht 199912000, 32p
plus appendix.
http://www.unibas.ch/vr-forschunglPERMOS,
PERMOS is part of the Global Terrestrial Network - Permfrost
GTN-P. I t is on track to be established and financed
MurWCorvattsch 2t1987
I l I I I I
I I I I l I
I I I , I I
""b-"..P"
I , ,
b m m O r t 4 A * m m b m m O r
m m m m m m m m m m ~ m m o m 2
Figure 2. Permafrost temperature at 11.6 m depth in borehole 211987 through MurteI-Corvatsch rock glacier. In the winters 198811989, 199511996
and 200112002 major snowfall came only in JanuarylFebruary, causing a strong cooling of the permafrost.
1 74

Methanogenesis under extreme .environmental conditions in permafrost
Soils: a model for exobiological processes
D. Wagner, S. Kobabe and E.". Pfeiffer
Alfred Wegener Institute for Polar and Marine Research, Research Unit Potsdam, Germany
INTRODUCTION
Permafrost within the solar system exists on seven planets
as well as several moons, which are characterized by cold
environmental conditions. Of special interest to present-day
research in astrobiology is the comparability of environmental
and climatic conditions of the early Mars and Earth. Martian
and terrestrial permafrost areas show similarmorphological
structures, suggesting that their development is based 011
comparable processes. New high-resolution images of the
Mars Orbiter Camera (MOC) imply water liquid in form or as
ground ice on Mars, which is an essential bio-physical requirement
for the survival of micro-organisms in permafrost.
Terrestrial permafrost, in which micro-organisms have survived
for several million years (Rivkina et al. 1998), is an
ideal model for comparative system studies regarding the
evolution, adaptation and survival strategy of microorganisms
under extreme conditions.
METHANOGENESIS
The microbial methane production (methanogenese) is the
terminal step during the decomposition of organic matter in
permafrost soils. Responsible for the methane production is a
small group of highly specialized microorganisms, called
methanogenic archaea, which are considered to be one of the
initial organisms from the beginning of life on Earth. They are
regarded as strictly anaerobic microbs, which can grow and
survive only under anoxic conditions, like in wet tundra environments.
Compared to Martian conditions, such a metab
lism would be favourable because the atmosphere on Mars
is dominated by C02 (95.3 %). Furthermore, they are characterized
by litho-autotrophic growth, whereby energy is
gained by the oxidation of hydrogen, and carbon dioxide can
be used as the only carbon source (Deppenmeier et al.
1996). Organic compounds, which were not verified an
Mars, are not required for growth. The special metabolism
qualities of the methane-forming archaea are the reason that
they are regarded as key organisms for further investigation
the adaptation strategies and long-term survival in extreme
environments such as those with terrestrial permafrost (Wagner
et al. 2001).
MICROBIAL METHANE PRODUCTION UNDER EX-
TREME CONDITIONS
The field investigations were carried out on the Samoylov island
(N 72", E 126") located in theLena DeltalSibena. The
study site represented an area of typical polygonal patterned
grounds with ice wedges. The permafrost soils were classified
as Glacic Aquifurbels and Typic Hisforthels with a
maximum thaw depth between 30 and 55 cm. The average
air temperature was -14.7 "C with a min. of -47.8 "C in
January and a max. of t18.3 "C in July.
The investigation of in situ methane production revealed
methane formation in the boundary layer to the permafrost at
temperatures between 0.6 and 1.2"C (Fig. 1) with a rate cf
about 1.0 nmol CH4 h-1 g-1. Without any additional substrate,
the methane production varied between 0.3 and 4.7 nmol
CH4 h-1 g-1 (Fig. 2). The highest activity could be determined
in the peat layer of the upper soil at an average temperature cf
10 "C. Nevertheless, the comparison of different vertical pr@
files to the methane production showed that the activity of the
micro organisms is independent of the soil temperature. After
addition of methanogenic substrates (acetate, Hz), the activity
drastically increased. The methane production in the peat
layer with H2 as substrate was about 2 times higher (13.0
nmol CH4 h-1 9-1) compared with acetate (7.2 nmol CH4 h-1
g-1) as substrate, while above the permafrost table activity the
was in the same order of magnitude (approx. 1.5 nmol CH4
h-1 g-1) with both substrates. Even the incubation of soil mate
rial from the active layer with H2 as a substrate showed a
significant methane production rate at -3 "C (0.1 - 11 -4 nmol
CH4 h-I g-I) and -6 "C (0.08 - 4.3 nmol CH4 h-1 g-I). The
existence of a methanogenic community, which has adapted
well to the low in situ temperature of permafrost soils is assumed
by our results.
In accordance with Panikov (1997), psychrophilic bacteria
have a significant part in the microbial community in cold
environments like permafrost soils. However, the isolation
175

and characterization of methanogenic archaea from permafrost
soils should clarify the possible growth characteristic (psychrophile
or psychrotroph) and the ecological significance d
these organisms.
80
A
CH
0 20 40 60 80 100 120 140 160 180 200
time Ihl
Figure 1. In situ methane production in the boundary layer
permafrost in the soil of the polygon depression.
-c H,
&acetate
-without
D 2 4 6 8 I 0 1 2 1 4
Cy production rate [nrnol h'h'q
to the
bial population will be influenced by the natural permafrost
system and by the interaction of the combined parameters.
REFERENCES
Deppenmeier, U., Muller, V., Gottschalk, G. 1996. Pathways of energy
conservation in methanogenic archaea. Archive of Microbiology
165: 149-1 63.
Panikov, N.S. 1997. A kinetic approach to microbial ecology in arctic
and boreal ecosystems in relation to global change, In W.C.
Oechel (ed), Global Change and Arctic Terrestrial Ecosystems
Berlin: Springer. 171 -188.
Rivkina, E., Gilichinsky, D., Wagener, S., Tiedje, T., McGrath, J. 1998.
Biogeochemical activity of anaerobic microorganisms from
buried permafrost sediments. Geomicrobiology 15: 187-193.
Wagner, D., Spieck, E., Bock, E., Pfeiffer, E". 2001. Microbial life in
terrestrial permafrost: methanogenesis and nitrification in Gelisols
as potentials for exobiological processes. In G. Horneck, C.
Baumstark-Khan (eds), Astrobiology: the quest for the conditions
of life Berlin: Springer. 143-159.
Wagner, D., Wille, C., Pfeiffer, E.". in preparation. Simulation of
annual freezing-thawing cycles in a permafrost microcosm for
assessing microbial methane production under extreme conditions.
Permafrost and Periglacial Processes, submitted,
Figure 2. Vertical profile of methane production rates without any
additional substrate as well as with HP and acetate in the soil of the
polygon depression.
First results concerning the diversity of methanogenic microflora
by fluorescence in situ hybridization (FISH) indicated
that Hz-using genera like Methanobacterium and Methanomicrobiurn
dominated over methylotrophic and acetotrophic
methanogenic archaea. Enrichment cultures, which were
dominated by the genera Methanosarcina, showed that these
organisms grow not only with the preferred substrate methanol
but also with C02 and HL as the only carbon and energy
source.
CONCLUSIONS AND FUTURE PERSPECTIVES
Methanogenic archaea, which can live solely on the basis of
water,minerals,carbondioxideandhydrogen, survived in
terrestrial permafrost for several millions of years. Since they
are still active, they had to develop different strategies to resist
extreme conditions like salt stress, physical damage by
ice crystals and background radiation. The presented results
demonstrate an essential basis for the understanding of the
origin of life. The studies can be used as a model for exobiological
processes to find traces of life in extraterrestrial habitats
and to look for possible protected niches e.g. on Mars.
Apart from the presented studies, simulation experiments
with permafrost microcosms can supply important results for
the understanding of microbial life under extreme conditions
(Wagner et al. 2003). Simulation of the freezing-thawing cycle
can help to understand the problem in which way the micro-
176

A physiognomic-based circumpolar Arctic vegetation map
D. A. Walker, M. K. Raynolds and, *N. G. Moskalenko
Alaska Geobotany Center, lnsfifute of Arctic Biology, University of Alaska Fairbanks
*Earth Ctyosphere Institute, Siberian Division, Russian Academy of Sciences, Russia
The Circumpolar Arctic Vegetation Map (CAVM) depicts
the distribution of 20 vegetation-physiognomy units in the
Arctic (CAVM Team 2003). The map is based on the principle
that environmental controls such as climate, topography,
soil chemistry and floristic provinces determine
which plants grow across the Arctic. The most important
environmental control of plant growth in the Arctic is summer
temperature. Temperature and vegetation data were
used to define bioclimatic subzones.
For purposes of display, the 20 units of the CAVM were
summarized into a much simpler map with 7 units (Table
l), based on dominant growth forms. The distribution of
these categories by bioclimatic subzone (Fig. 1) is shown
in Table 2.
Table 1. Condensed physiognomic legend with CAVM units.
Glaciers: 17. Nunatak area; 18. Glacier
Barren, herbs: 1. Cryptogam, cushion forb barren;
2. Cryptogam barren (bedrock); 3. Noncarbonate
mountain complex; 4. Carbonate
mountain complex
Graminoid, prostrate shrubs: 5. Rushlgrass, forb
cryptogam tundra; 6. Grarninoid, prostrate dwarf
shrub, forb tundra
Prostrate, hemi prostrate shrubs: 9. Sedge/grass,
moss wetland; 13. Prostrate dwarf shrub, herb
tundra; 14. Prostratdhemiprostrate dwarf shrub
tundra
Graminoid, dwarf shrubs: 7. Non-tussock sedge,
dwarf-shrub, moss tundra; 8. Tussock sedge,
dwarf shrub moss tundra
Erect dwarf shrubs, low shrubs: 10. Sedge, moss,
dwarf shrub wetland; 11. Sedge, moss low shrub
wetland; 15. Erect dwarf shrub tundra; 16. Low
shrub tundra
Water: 19. Lakes. 20. Ocean
Figure 1. Bioclimate subzones of the circumpolar Arctic.
The glacier category excludes the Greenland Ice Sheet
(1697 km*), which spans numerous bioclimatic subzones.
Glaciers cover the largest area in Subzone C, but cover
the greatest percentage of Subzone A. Barren vegetation
types are common in Subzones A and B, especially on the
calcareous soils of the Canadian Arctic Islands. Barren
areas in Subzone C are mainly on the Canadian Shield
and the Taymyr Mountains; and in Subzones D and E, are
due to the mountains of Chukotka and Alaska.
Graminoid-dominated vegetation types with prostrate
shrubs are found in the southern Canadian Arctic Islands
(Subzones B and C). Graminoid types with erect shrubs
(including tussock tundra) are more common in Alaska
and Yakutia (Subzones D and E).
Prostrate-shrub dominated vegetation types are common
on the southern Canadian Arctic Islands, northern
mainland Canada and the Taymyr Peninsula (mainly Subzone
C). Erect shrubs are found in western Russia,
southern Chukotka, southern Keewatin and the Ungava
Peninsula (mainly Subzone E).
177

~
~
~ prostrate
I
Vegetation Physiognomy
C"" 1
Water and non-arctic
Q Glaciers
Lz] Barren, herbs
" "_ I_
Gramhold,
shrubs
Prostrate, hemiprostrate
shrubs
Grarninoid,
dwarf shrubs
Erect dwarf shrubs,
low shrubs
.. .
Figure 2. Distribution of vegetation physiognomy in the circumpolar Arctic
Table 2. Area of physiognomic categories in bioclimate subzones
(1000 km2).
T - Subzo -
Physiognomic Unit - A - B - c -
11 E Totul
Glaciers
B,men, herbs
Graminoid, prostrate
shrubs
Prostrate, hemiprostrate
shruhs
Graminoid, dwarf
shrubs
Erect dwarf shrubs,
95
61
2
37
0
63
242
71
130
0
127
332
375
305
34
74
343
162
19
609
2
23 I
0
68
505
360
1208
610
560
1148
low shrubs
0 0 - 0 - 272 - 924 1196
Total
194 506 1174 1479 1729 SO83
---
1
zone D is mostly graminoid, dwarf shrubs. Subzone E is a
mix of erect shrubs and graminoid, dwarf shrubs.
ACKNOWLEDGEMENTS
This research project is funded by National Science Foundation
Grant OPP-9909929, with support from the US Fish
and Wildlife Service and the Circumpolar Arctic Flora and
Fauna Project.
REFERENCES
Subzone A is predominantly glaciers and barren. Sub- Circumpolar Arctic Vegetation Mapping Team, in press. Circumpolar
zone B is predominantly barren and prostrate shrubs.
Arctic vegetation. Conservation of Arctic Flora and Fauna (CAFF)
Map No. 1.
Subzone C is a mix of barren areas (Canadian Shield);
Scale 1 :7,500,000. US Fish and Wildlife Service, Anchorage,
AK.
graminoid, prostrate shrubs; and prostrate shrubs. Sub-
178

Permafrost and the Colville River Delta, Alaska
H. J. Walker
Department of Geography and Anthropology, Louisiana State University, Baton Rouge, LA 70803-4105, USA
Although the study of frozen ground has a fairly long history,
it was not until 1943 that Siemon Muller first used the
word permafrost in a classified volume for the US. Army
entitled Permafrost or Permanently Frozen Ground and
Related Engineering Problems. In 1947, after declassification,
this monograph was published as a hard-cover
book and then became available to the public (Muller
1947). It can be surmised that Professor Muller, in his
Stanford University office, never anticipated that within 20
years there would be an international conference on the
subject and that an international permafrost association
would soon follow.
Research on the geomorphology of the Colville River
delta, which was begun in 1961, included permafrost as a
main item of concern. At the First International Conference
on Permafrost, held at Purdue University, USA in 1963
(i.e., 40 years ago), a paper on the effect of permafrost
and ice wedges on riverbank erosion was presented
(Walker and
Arnborg 1966). Measurements collected included the
depth (horizontal) of thermo-erosional niches (a term borrowed
from the Russian termerosionaja niza) in the permafrost-bound
banks of the delta’s distributaries. Additional
surveys were made of the included ice wedges
along which block collapse occurred (Fig. 1) and benchmarks
were established to monitor thaw rate of bank retreat.
These stations were occupied periodically for the
next 30 years. A number of seasons of measurements
show that the rate of bank retreat varies with bank composition,
bank height, and exposure to erosional processes
(Walker 1983).
Lakes within the delta are highly variable in size,
shape, and origin. All are affected by permafrost to some
extent. Those sufficiently deep have a talik beneath but
most are shallow and floored by permafrost with only a
thin active layer developing during the summer. The most
numerous bodies of water are those present in Iowcentered
polygons bordered by ice wedges (Fig. 2).
Possibly the most unique of the lakes affected by permafrost
in the delta are those perched in the sand dunes.
Perched lakes can only continue to exist so long as there
is an aquaclude between the lake’s water table and the
regional water table; in the case of the Arctic this aquaclude
is often permafrost. Drainage of these ponds occurs
by overflow as the accumulated snow melts and by flow
through the surrounding sand as the active layer develops.
The maximum thickness of the active layer within the
dunes of the Colville delta by the end of the summer is
generally less than two metres. Thus, those ponds in interdune
depressions and in blowouts that are more than two
metres deep retain water throughout the year (Walker
2002).
Morphologically, lake tapping provides some of the
most sudden and drastic changes in the delta. Both lake
expansion and river-channel migration lead to tapping
which usually occurs between polygons because the ice
wedges thaw faster than the surrounding permafrost.
Once tapping occurs, lakes fluctuate in level with stage. At
179

low stage lake depths are reduced sufficiently so that Walker, H. J. 2002. Landform development in an arctic delta; the roles
permafrost begins to form in bottom materials. They also of snow, ice and permafrost. In M. K. Hewitt et al. (eds), Landscapes
of Transition, Dordrecht: Kluwer Publishers: 159-183.
become subject to deposition during the river‘s flood stage
and deltas form at the lake’s entrance.
Sixty years ago, Muller emphasized the engineering
problems associated with permafrost. During most of the
subsequent period of time, most of the permafrost-related
research in the Colville delta was basic. However, in the
199Os, with the exploration for and development of the oil
industry in the delta, research became more applied. It
was aimed at providing ‘I... information essential for engineering
design and evolution of potential environmental
impacts ...” (Jorgenson et al. 1998). Such research, which
is being conducted by a variety of research teams, is providing
much more detailed information about the role of
permafrost in the formation and modification of delta
forms.
REFERENCES
Jorgenson, M. T., Shur, Y. L., and Walker, H. J. 1998. Evolution of a
permafrost-dominated landscape on the Colville River delta,
Northern Alaska. Proceedings: Seventh International Permafrost
Conference, Centre d’etudes nordiques, Universite Laval: 523-
529.
Muller, S. W. 1947. Permafrost or Permanently Frozen Ground and
Related Engineering Problems. Ann Arbor: J. W. Edwards Press.
Walker, H. J. and Arnborg, L. 1966. Permafrost and ice-wedge effect
on riverbank erosion. Proceedings: First International Permafrost
Conference, National Academy of Science, National Research
Council, Washington D. C, Publication 1287: 164-171.
Walker, H. J. 1983. Erosion in a permafrost-dominated delta. Proceedings:
Fourth International Permafrost Conference. National
Academy of Science Press, Washington D. C.: 1344-1349.
180

Establishing four sites for measuring costal cliff erosion by means of
terrestrial photogrammetry in the Kongsfjorden area, Svalbard
B. Wangensteen, T. Eiken, J. 1. Solid, *R. S. 0degArd
Department of Physical Geography, University of Oslo, Norway
*Gjovik University College, Norway
The Kongsfjorden area is situated at the western coast of
the Svalbard archipelago (Fig. I). The area has continuous
permafrost with measured depths around 140 meters
near the shores of Ny-Alesund (Liestarl 1976). The mean
annual air temperature in Ny-Alesund is -5.8" C (1975 to
2000) (KetiI Isaksen, pers. comm.). The tidal range is
about 2 meters. The shores of the area are a mixture of
cliffs in unconsolidated material, cliffs in bedrock and glacier
fronts terminating in the sea. The selected four sites
consist of low cliffs in bedrock and unconsolidated material,
both with stony beaches in front (0deg5rd et al.
1987). The geology at three of the sites is lime- and
dolostone (Carbon and Perm) while one site is in an area
of gneiss (upper Proterozoic). The unconsolidated beach
material is of Quaternary origin (Hjelle 1993). The area
has earlier been subjected to some related research activities
concerning coastal processes e.g., cryogenic
weathering of cliffs (0degard and Sollid 1993).
At all sites a fixed point for global reference was established;
a bolt was drilled into the bedrock and its position
measured by GPS. The positions of the camera stations
were measured by GPS and surveyed from the fixed point
to create both a global and a local reference. The photos
were acquired with a stereo overlap using a Hasselblad
camera with a 60 mm lens. At two of the sites bolts to be
used as photogrammetric control points were drilled into
the cliff wall as well. At the sites consisting of
unconsolidated material no control points were established.
The distance between the camera stations and the
cliff wall varied between 7 and 15 meters for the four sites.
The distance between the camera stations was approximately
half of the photo distance. The coordinate system
is transformed in a manner to simulate the geometry of
aerial photogrammetry. The z-axis is pointing out of the
cliff and the xy-plane is parallel to the cliff wall and also
the line between the photo stations. The camera was
mounted on top of the theodolite with a special device
making it possible to measure the exact position of the
camera and to ensure that the photo stations are on line
parallel to the cliff wall. The camera positions are also
used in the photogrammetric orientation of the scenes. For
the scenes from the sites of unconsoli dated material this
set-up also gives the opportunity to measure the photo
Figure 1. Map showing location of the investigated area.
angle used for exterior orientation of the scenes. This is
important since no control points are available. The photos
were scanned at a resolution of 11 pm (2,300 dpi) and
imported into the Zll Imaging digital photogrammetric
workstation software. To be able to use the Zll Imaging
software the coordinate system had to be rotated, as explained
earlier, and all coordinates up-scaled by a magnitude
of 100, both to simulated aerial photography. After
the photo pair or triples have been used to create a photogrammetric
stereo model it is possible to automatically
generate a high-precision digital terrain model (DTM). The
Match-T algorithm in Zll Imaging utilizes cross-correlation
to detect the same points in the different photos of the stereo
model and then measures the elevation by measuring
the y-parallax. Photo distances in the range of 7 to 15 m
give a scale of 1 :I 17 to 1 :250. With the above-mentioned
scanning resolution the XY-resolution is in the order of 1.2
to 2.5 mm. The accuracy of elevation measurements done
by photogrammetry is about 0.20 to 0.40%0 of the photo
distance. This gives a theoretical elevation accuracy (Zdirection)
of 1.4 to 2.8 mm for the 7 m photo distance and
3.0 to 6.0 mm for the 15 m photo distance.
-4
181

The automatically-generated DTM for one of the sites
is shown in Figure 2 as contour lines on top of an orthophoto
of the cliff wall. The height and width of the area
covered by the model is 3.3 m and 2.8 m respectively.
When investigated in the digital photogrammetric workstation,
the terrain model follows the terrain quite nicely. Precision
estimates within the Zll Imaging software gives a
value of 1 to 2 mm for the Z-accuracy at this site. The
photo distance used here was 7 meters.
REFERENCES
And& M.-F. 1997. Holocene Rockwall Retreat in Svalbard: A triplerate
evolution. Earth Surface and Processes and Landforms 22:
423-440.
Hjelle, A. 1993. Geology of Svalbard. Polarhhdbok No. 7, Norsk Polarinstitutt,
Oslo.
Liestol, 0. 1976. Pingos, springs and permafrost in Spitsbergen. In
Norsk Polarinstitutts Arbok 19757-29.
0degArd, R., Sollid, J.L. and Trollvik, J.A. 1987. Coastal Map Svalbard
A 3 Forlandssundet 1:200.000. Department of Physical Geography,
University of Oslo.
0degArd, R. Sollid, J.L. 1993. Coastal cliff temperatures related to the
potential for cryogenic weathering processes, western Spitsbergen,
Svalbard. Polar Research 12(1): 95-106.
Fj~ur~ 2. ~ont~ur lines of 5 crn
~ha~a. The size of the rno~~i ar
~ i ~ ~ ~ ~ h ~ ~ ~ ~ ~ i ~ ~ l ~ ~ .
The method of close up digital terrestrial photogrammetry
seems suitable for creating accurate digital terrain
models of coastal cliffs and hence promising for calculating
erosion rates in the order of mmlyear. The intention is
to acquire a new set of photos of the same sites in 2004 to
generate new terrain models that will enable the erosion
rate calculations. Andre (1997) reports a rock wall retreat
in the range 0 to 1.58 mmlyear in the same area. The
coastal cliff erosion rate is unknown but believed to be
higher than the rock wall retreat rate and hopefully greater
than the accuracy of the technique used.
The fieldwork was financially supported by the INTASproject
“Arctic coasts of Eurasia: dynamics, sediment
budget and carbon flux in connection with permafrost
degradation’’ (INTAS-2001-2329) and the Norwegian
Research Council on behalf of the Norwegian Polar
Committee.
182

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Av~KI~~Ie Information
Laboratory experiments towards better understanding of surface buckling
of rock glaciers
M. Weber and A. Kaab
Glaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland
Wahrhaftig and Cox (1959) assumed permafrost creep
to be the main transfer process in rock glacier flow. A
number of boreholes on active rock glaciers give insight
into material properties, rheology and the thermal conditions
involved (Haeberli et al. 1998). Information on
kinematics of rock glaciers has been obtained from terrestrial
measurements and computer-aided aerial photogrammetry
(Kaab 1998). Numerical modelling showed
that transverse ridges might develop after a change in
slope if two layers with different viscosity exist (Loewenherz
et al. 1989).Yet still many questions concerning the
interaction between kinematics and structure of a rock
glacier remain unanswered. The characteristic topographic
surface structures on rock glaciers, namely transverse
ridges, can be regarded as a physical expression
of these interactions. Within the framework of a diploma
thesis aiming to analyze the occurrence of surface
structures in combination with surface parameters, a
creeping mass has been physically modelled in a laboratory
experiment to qualitatively estimate possible
processes involved in such surface buckling. The study
addresses the following questions: is there a suitable
substitute for the debris-ice mixture of a real rock glacier
to be used in laboratory simulations Is a decrease in
longitudinal slope sufficient to trigger the evolution of
transverse ridges in a viscous material If not, which
material properties allow transverse ridges to develop
In our experiments the food additive Xanthan Gum
(E415) has been tested. Xanthan gum is a microbial
polymer prepared commercially in a pure culture fermentation
of the bacterium Xanthomonas campestris.
The molecular structure of Xanthan Gum results in
pseudo plastic flow behaviour. Xanthan Gum is used in
food engineering as well as cosmetics, and the petrochemical
industry in order to control flow processes. Our
experiments were carried out on a ramp built from perspex.
The ramp is constructed such that its slope can be
varied. (The discontinuity between the ramp and the
horizontal base is referred to here as the "break of
slope"). A container is fixed to the top of the ramp, used
to provide equal starting conditions for each experiment.
In changing the mixture of two compounds differing in
viscosity and in solids content such as sand andlor
gravel, the rheological behaviour of the material has
been varied in a systematic way.
In summary, the experiments showed the following
trends:
- No or only very minor transverse ridges at or near the
break of slope are created if the material was well
mixed and "homogeneous" (Le., no solids content).
- If two components of different viscosities but no solids
content are only slightly mixed ("heterogeneous"),
fine transverse ridges could develop near the break
of slope. After sufficient material has crept downwards
and starts to accumulate in the horizontal,
ridges develop above the break of slope.
- Ridges also develop if sand andlor gravel are added
to a single-component mass.
- If a mass comprised of sand- and gravel-containing
parts of different viscosity is only slightly mixed or
applied in two separate layers, distinctive ridges appear.
- In another experiment, where a slightly mixed mass
with sand and gravel ("heterogeneous") formed the
bottom layer covered by a dry gravel brittle top layer,
distinctive ridges developed. Moreover, due to the
compression at the point of break in slope, single
gravel particles were dislocated and fell into the
newly-formed furrows. Viscous material was pushed
from underneath to the surface forming ridges of increasing
amplitude.
The findings from these experiments exhibit analogies
to real rock glaciers. Although the observed processes
are certainly not equivalent to those occurring in
nature, the experiments yield structures comparable to
those of real rock glaciers. A development model for
transverse ridges can be deduced from the experiment
described above in which a brittle layer has been forced
up through viscous material. According to this model, it
may be possible that transverse ridges on rock glaciers
can develop due to the compressing flow of the viscous
ice-debris-mass and its interaction with an overlying brittle
layer (i.e., the active layer). It may also be possible

due to the entire flow process. In any case, a heterogenous
variation of viscosities, or solids content, respectively,
seems to favour or even be a prerequisite for
transverse surface structures to develop. Such conceptual
models are presently being investigated in relevant
high-precision field surveys.
ACKNOWLEDGEMENTS
Special thanks to STAERKLE & NAGLER AG, Zollikon,
Switzerland for providing Xanthan Gum.
Weber, M. et al.: Surface buckling of rack glaciers,..
REFERENCES
Haeberli, W. 1985. Creep of mountain permafrost: internal structure
and flow of alpine rock glaciers. Mitt. Versuchsanst. Wasserbau
Hydrologie Glaziologie ETH Zurich. 77 pp.
Kaab, A., Gudmundsson, G.H. and Hoelzle, M. 1998: Surface deformation
of creeping mountain permafrost. Photogrammetric
investigations on Murtel Rock Glacier, Swiss Alps. Proceedings
of the Seventh International Conference on Permafrost, Yellowknife,
Canada, Collection Nordicana 57: 531-537.
Loewenhen, D. S., Lawrence, C. J., Weaver, R. L. 1989: On the
development of transverse ridges on rock glaciers. Journal of
Glaciology 35 (121): 383-391,
Wahrhaftig, C., Cox, A. 1959: Rock Glaciers in the Alaska Range.
Bulletin of the Geological Society of America 70: 383-436.
Weber, M. 2003: Oberflachenstrukturen auf Blockgletschern. Diploma
thesis, University of Zurich.
184

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Russian Information Transfer Programme
P. J. Williams and 1. M. T. Warren
Scott Polar Research Institute, Cambridge, United Kingdom
The longest tradition of permafrost knowledge and the
largest body of accumulated information is Russian. Russian
studies can be traced back centuries, while it was
only in the second world war that permafrost became of
broad practical interest and an object of much scientific investigation
in North America. In Finland, Norway and Sweden
with their cold climates, and elsewhere, occasional
studies have been made over two centuries at least. But
apart from the contributions of several key figures, for example
Taber, Beskow or Troll, only in Russia were there
large-scale organised studies of frozen ground and permafrost
and the corresponding technologies, through the
first half of the twentieth century.
This Russian contribution continues and will increase,
considering how many people live in a permafrost environment
there, the extent of industrial (oil, gas, mineral
development) and other human activities (including those
of many native peoples). Development of Alaska and its
natural resources and of northern Canada has brought
similar intense interest in the scientific, industrial, geotechnical
and general environmental conditions of the
permafrost regions there. Such is modern society, permafrost
is of global significance.
Most permafrost literature is in Russian or English, both
major languages but not, for most individuals, interchangeable.
The difficulties concern engineers, financiers,
politicians and many others now involved with the cold regions.
Accessibility to the information, however, is of the
utmost importance - it may be extremely costly in financial
and in human terms, to remain unaware of the hard-won
expertise of others.
The Russian Information Transfer Programme of the
Scott Polar Research Institute in Cambridge, England, is
designed to bring the significance of ‘permafrost‘ literature
in the Russian language to those who cannot read it. It
goes further than translation: it is a scholarly undertaking
to ensure that the Russian work is as meaningful to English-
language readers as to the Russian authors themselves.
The history of the twentieth century was such that
studies relating to Siberia, to the oil and gas industry, to
185
military interests, and to the cold regions generally were
not widely revealed to those outside the Soviet Union.
Even more important was that the general lack of scientific
exchange resulted in the development of concepts and
models distinctly Russian on the one hand and distinctly
‘English language’ or ‘Western’ on the other. And however
etymologically-polished a translation may appear, it loses
much of its value if it is not quite understood by the specialist
reader. Think merely of ‘cryolithological regions’,
‘massive permafrost’, ‘basal ice’- terms quite acceptable to
a first-rate translator but full of uncertainty to the permafrost-sawy
English language reader.
The first undertaking under the nascent Programme
was Professor Yershov’s Obshchaya Geokriologiya -
General Geocryology (1998), which is a comprehensive
overview of Russian permafrost science and technology.
The aim was to publish the book to a comparable standard
to that if had been authored in English. This was a
challenge but, even if we were not completely successful
in meeting it, the book has given to English language
readers a unique insight into the depth of Russian knowledge
and its implications.
Building on this experience, an English-language folio
volume was prepared to accompany the original maps of
the 16-sheet highly informative Geocryological Map of
Russia and Neighbouring Republics. All legends are
translated, with additional explanations and a glossary.
The Map set has immediate practical applications in planning,
and for major engineering and other works. It gives
insight into permafrost distribution and origin, active layer
conditions and many other features, as well as into the
manner in which such information is interpreted and utilised.
The second edition (Williams and Warren 2003) of
this folio volume (together with the maps) is being published
shortly to meet the increased global interest in oil
and gas especially, and in other aspects of the Russian
cold regions. It includes refinements of the glossary, the
translations and other components of the volume. Both
Yershov’s book and the Map have led to extensive discussions
of interpretation with numerous specialists.

Williams, P. J. et al.: Russian Information Transfer Prograrnrne
The next major undertaking has been the preparation
of a volume of selected articles from Kriosfera Zemli -
Earth’s Cryosphere, the leading Russian journal devoted
to permafrost and related topics. This work too has been
carried out as a scholarly undertaking and, because of the
diversity and importance of topics covered (for example,
electrical and mechanical properties of frozen ground;
permafrost distribution in coming years; gas hydrate stability;
geotechnical foundation techniques; biological activity
in permafrost) the advice of even more specialists
(some never before involved with permafrost), has been
sought. The aim is to continue to publish an English language
version of this unique journal - but this will depend
on the degree of financial support achieved through sales
and other sources.
The Programme has benefited from a generous, collaborative
approach from both the Russian and Englishlanguage
sides, and especially from the authors, editors
and publishers of the Russian studies. The Programme
has led to new linkages and has given those working with
it a fuller bibliographic insight, as well as a depth of understanding
of the people and organisations that are contributing
to our knowledge of geocryology. Consequently the
Programme is also able to make other contributions in the
form of advice, and, when funds are available, of literature
searches and analyses and ‘custom’ translations. European
Union funds have been awarded for visits to meetings
by Russian specialists, especially relating to pollution
in permafrost regions, strengthening the Programme in
this topic.
Political developments and the significance of the permafrost
regions point to a growth in coming years of the
Russian Information Transfer Programme, but this will depend
on the response of those who utilise it. Most agencies
funding research doubtless see these activities as
being peripheral to their responsibilities. Yet it is cheaper
to benefit by the experience of others than to rediscover
the problems and the expense of uncertain solutions, especially
when the costs of making such knowledge available
are small indeed compared to the financial and other
dangers of proceeding in ignorance.
i
REFERENCES
Yershov, E. 1998. General Geocryology: (originally published as
Obshchaya Geokriologiya, Nedra, Moscow). Cambridge University
Press, 580pp ISBN 0 521 47334 9.
Williams, P.J. and Warren I.M.T. 2003.The English Version of the
Geocryological Map of Russia and Neighbouring Republics.2nd.
Edn. Original maps, guide, folio, I-X, & 21pp. Cambridge University,
Carleton University, University of Moscow. Ottawa: Collaborative
Map Project. ISBN 0-9685013-0-3.
186

8th international Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Permafrost as a natural cryobank of late Pleistocene and modern plant
germplasm
Institute of Cell Biophysics, Russian Academy
S. G. Yashina, E .N .Gakhova and *S. V. Gubin
of Sciences, Pushchino, Moscow Region, Russia
"Institute of Physicochemical and Biological Problems in Soil Science, Russian Academy of Sciences, Pushchino, Moscow
Region, Russia
INTRODUCTION
RESULTS AND
DISCUSSION
Conserving the world's biological diversity has recently become
a global problem. The scientific community is now
ready to use the advances in cryobiology to initiate long-term
strategic and sustainable genetic resource cryobanks for
saving endangered species, in addition to the traditional
means such as nature reservation, botanical gardens, etc.
(Gakhova 1998, Tikhonova and Yashina 1998). In recent
years, the viable spores of mosses (Gilichinsky et al. 2001)
and seed embryo cells of the higher plants (Yashina et al.
2002) have been discovered within Siberian late Pleistocene
permafrost. It was demonstrated that they were able to de
velop and grow. In connection with this, new prospects have
been revealed for cryopreservation and long-term banking
germplasm of the modern wild plant. In this study we show
current research on vital functions of higher plant seeds from
permafrost deposits and present some data for the use d
permafrost for the conservation of modern plant seeds.
MATERIALS AND METHODS
The seeds and unripe fruits were evoked from the fossil barrows
of ground squirrel in the late Pleistocene ice-loess
sediments of North Yakutiya (Gubin and Khasanov 1996).
The burrows were found in the constant annual subzero temperature
layer at a depth of IOto 18 m. Radiocarbon age d
the paleontological material was determined to be an average
of 30 thousand years. Seeds were identified by means of a
collection of plant seeds from the Kolymsko-lndigiskaya
lowland (Yashina et al. 2002). The fossil seeds were into
duced into in vitro culture on an agar nutritious medium to ascertain
their vital capacity. Methods of plant tissue culture and
clonal micropropagation of plant in vitro, light microscopy and
microphoto were used.
The long-term experiments on storage of modern wild
plant (14 species) seeds were performed in a permafrost
zone at various temperature conditions. Viability of modem
plant seeds was determined by periodic control over the lab
ratory germinating capacity.
Seeds of sufficient diversity were found in the burrow studied
(up to IOspecies). Attempts at seed germination were undertaken
for each species, but physiological viability of seeds
was discovered only in 3 species: stenophyllous campion
(Silene stenophylla Ledeb.) of the pink family (Caryophyllaceae)
and 2 species, identified accurate to the genus: arctous
(Arcfous spp.) of the heather family (Ericaceae) and
persicaria (Polygonum spp.) of the persicaria family (Polygonaceae).
The plantlet of persicaria seed has germinated in vifro
culture to cotyledon leaf stage (Fig. I). Callusing from eE
dospem cells of arctous seeds has been obtained also in
culture. It tumed out that campion seeds are more tolerant b
long-term storage at a subzero temperature. The initial stage d
embryonal root growth and callus formation from the resulting
roots cells has been induced on several of campion seeds.
Development of embryonal roots has not occurred because d
degradation of the storage tissue. At the same time, fragments
of embryonal tissue in unripe fruits of campion have been
shown to have the properties of enlargement (Fig. 2) and organogenesis
formation. At present, the plant material obtained
provides the basis for clonal micropropagation and the subseguent
recovery of plants of full value. The long-standing experiments
on storage of wild plant seeds are performed in a
permafrost zone at different temperature regimes in order b
use permafrost for conservation of the plant genetic resources.
Part of the seed samples were placed into the glacier
(Kolymsko-lndigiskaya lowland) at mean annual temperatures
of -15OC, the rest were placed into an underground
depository in the Melnikov permafrost Institute of SD RAS
(Yakutsk) and stored at -2,7OC. Research on viability of
seeds by determination of laboratory germination capacity
over 1.5, 2 and 5 years have shown that glacier conditions
had some advantages over the underground storage d
seeds. At thesametimetheseedsofsomespecies had
more high germination capacity in underground storage in relation
to control (laboratory storage at +4OC). It is interesting
that microbiotic seeds (Anemone sylvestris L., Pulsatilla patens
(L.) Mill.) preserved high viability in the glacier as well
as in the underground depository, whereas these seeds rag
idly lost germination capacity at +4OC. Both easily germinat-
187

ing seeds (Dianthus fischeri Spreng, D. superbus L., Leucanthemum
vulgare Lam.) and seeds with biological dormancy
(Liliurn martagon L., Lavafera thuringiaca L.) had high
viability after storage in the glacier. Seeds of several species
in the pink family demonstrated especially high viability after
storage in the glacier for 5 years. It is noteworthy that the viable
fossil stenophyllous campion belongs to this same family.
Yashina, S. G. et al.: Permafrost as a natural cryobank ...
ACKNOWLEDMENT
This work was supported by the Russian Foundation of Basic
Research (grants 99-0449369, 01-0449087 and 0244-
49369).
REFERENCES
Fi~ure 1. The ~lantlet of fossil ~ ~~si~aria seed
Gakhova, E.N. 1998. Genetic cryobanks for conservation of biodiversity.
The development and current status of this problem in
Russia. Cryo-Letters, suppl.1: 57-64.
Gilichinsky, D.A. et al 2001. How long the life might be preserved
In (Giovannelli F., ed.), Proceedings of the International Workshop
((The Bridge between the Big Bang and Biology)), Consiglio
Nazionale delle Ricerche of Italy: 362-369.
Gubin, S.V. and Khasanov, B.R. 1996. The fossil barrows of Mammalia
from permafrost sediments of Kolymsko-lndigirskaya
Lowland. Dokl.Akad.Nauk 346(2): 278-279 (in Russian).
Tikhonova, V.L. and Yashina, S.G. Long-term storage of endangered
wild plant seeds. Physiol. Gen. Biol. Rew. 13(1): 1-33.
Yashina, S.G. et al. 2002. Viability of higher plant seeds of late
Pleistocene age found in permafrost deposits as determined by
in vitro culturing. Dokl.Akad.Nauk 383(5): 714-717 (in Russian).
Figure 2. The fossil campion seed with fragment
bryonal tissue
of enlarging em-
CONCLUSION
For the first time, it has been demonstrated that cells of fragments
of unripe fruits of stenophyllous campion and seed embryo
cells of this species, as well as seed embryo cells d
two unidentified species of the genera arctous and polygonum,
preserved viability during thirty thousand years in
permafrost. Physiological activity of seed cells has been discovered
by the in vitro cultural method. This pioneering
method for the conservation of seeds of modern wild species
under permafrost conditions may be promising because this
type of permafrost storage is protected from natural disasters
and anthropogenic effects, is relatively inexpensive, and can
maintain a stable subzero temperature. Many importance
questions remain to be resolved including the identification d
natural mechanisms of adaptation that allow plant cells and
tissues to preserve vital capacity for thousands of years at
subzero temperatures. Also, there is a need to study tern
perature regimes and other physicochemical conditions the for
long-term storage of plant germplasm in permafrost.
188

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Active layer monitoring in Northeast Russia: regional and interannual
variability
D. Zamolodchikov, *A. Kofov, **D. Karelin and ***V. Rarzhivin
Centre for Ecology and Productivity of Forests, Russian Academy of Sciences, Moscow, Russia
The CALM network (Brown et al. 2000) presently includes
three sites in northeast Russia (Chukotka district). The
first site, Cape Rogozhny, was established in I994 on the
northern coast of Onemen Bay (64"49'N, 176'50'E) of the
Bering Sea. The second site, Mount Dionisy, started in
I996 and is located about 25 km south of Anadyr city
(64"34'N, 177'12'E). Both sites are dominated by cottongrass
tussocks on Gleyi-Histic Cryosols. Frost cracks and
frost boils are well developed. The third site, Lavrentia,
was initiated in 2000 in the vicinity of Lavrentia settlement
(65036'N, 171°03'W). The site consists of wet sedge-Salix
mosses tundra without developed microrelief on Gleyi-
Histic Cryosols. This report was developed as a result of
the CALM synthesis workshop held in November 2002 at
the University of Delaware.
Grids (100x100 m) are used for active layer monitoring
at the above sites. Thaw depth is measured using a steel
rod (121 points per grid). The surface soil moisture is
monitored at the Lavrentia site with a Vitel hydra probe
(Hydra soil moisture probe 1994). Only end-of-season
data are available for Cape Rogozhny and Mount Dionisy.
Seasonal measurements were done at the Lavrentia site
(3-4 point per summer).
The climate conditions at Cape Rogozhny and Mount
Dionisy are well described by data from the Anadyr
weather station (64"47'N, 177'34'E). The closest weather
station to the Lavrentia site is at the Uelen settlement
(6601 O'N , 169050'W).
The mean annual air temperature at the Anadyr and
Uelen stations weather stations are fairly similar (Fig. 1A)
in absolute values and interannual dynamics (R=0.89).
Some temperature increase is observed from I999 to
2002, but the warmest period was 1995 to 1997. For the
period of CALM observations (1994-2002), a trend in the
mean annual temperature is not noticeable.
An important climatic predictor of seasonal thawing is
thawing degree days (DDT) value (Brown et al. 2000). The
DDT values are different in the Anadyr and Uelen stations
(Fig. IB). The difference is explained by colder summers
in the more marine conditions of Uelen and Lavrentia locations.
The warmer winters at the above locations lead to
similarities in the annual temperature regimes.
*Scientific Center "Chukofka': Anadyr, Russia
**Biological department, Moscow State University, Moscow, Russia
***Komarov Botanical Institute, St-Petersburg, Russia
1994 1995 1996 1997 1998 1999 2000 2001 2002
200 I
1994 1995 1996 1997 1998 2000 1999 2001 2002
3s 1
_I" 1
- - - L, - - Cane Ronozhnv
"a- Mount of Dionisv
1994 1995 1996 1997 1998 1999 2000 2001 2002
Year
Figure 1. Dynamics of (A) mean annual air temperature, (B) DDT, and
(C) end-of-season thaw depth at the CALM sites in northeast Russia.
The mean end-of-season thaw depth in Cape Rogozhny
is 44 cm with interannual variations from 38 to 50 cm
(Fig. IC). The peak of seasonal thawing was registered in
1996 and 1997 (48 and 50 cm respectively).
The mean end-of-season thaw depth at Mount Dionisy
is 49 cm with interannual variations of 45 to 53 cm. The
most northern CALM site (Lavrentia) is characterized by
the largest end-of-season thaw depth (60 cm), with small
interannual variations (59 to 61 cm).
End-of-season thaw depth at Cape Rogozhny is correlated
(R=0.62) with the square root of DDT (Fig. 2). The
correlation is more pronounced (R=0.71) for thaw depth
189
b-*k

and mean annual temperature. This result stresses the
role that warm winters play in determining frozen ground
temperatures and corresponding lower energy requirements
for the thawing season.
The correlation of end-of-season thaw depth and
DDTo.5 is poor at the Mount Dionisy site (Fig. 2). This has
two reasonable explanations. The first is related to the
early dates of measurements (near 10 August) in some
years (1 997, 2000). Actual end-of-season thaw depth can
be different from the measured values in these years.
The second is due to landscape position of the site where
the underground water flows determine the spatial patterns
of soil thawing. Under these conditions the actual
thaw depth is determined more by summer precipitation
than by air temperatures.
The end-of-season thaw depth for the Lavrentia site is
well-correlated with DDT0.5 (R=0.96). However, the time
series are limited for the site (n=3).
Analysis of the current CALM grid data allows us to
characterize the main influences of spatial variability on
thaw depth. At the Cape Rogozhny site the above factors
are disturbances of vegetation and soil cover (frozen boils,
tracks of all-terrain vehicles). In the Mountain Dionisy site,
the spatial variability of the active layer is determined by
subsurface water flows, corresponding to depressions of
mesorelief. Frost boils also input to active layer variability
at this site. The main control factor of end-of-season thaw
depth at the Lavrentia site is the thickness of the soil organic
layer.
Seasonal pattern of measurements at the Lavrentia site
make it possible to reveal the changes of spatial variability
control factor during soil thawing. At the beginning of the
warm season, the thickness of organic soil layer, surface
soil moisture and thickness of moss cover demonstrate
approximately the same relation to thaw depth (absolute
values of R coefficients vary from 0.3 to 0.4). In the course
of the season the role of organic layer increases (R
achieves -0.6), while the input of other factors decreases
(absolute value of R diminishes to 0.2).
The present review of regional CALM studies in northeast
Russia results in the following conclusions:
there are no evident trends in end-of-season thaw
depth in northeast Russia;
drainage ways, surface disturbances and the thickness
of the organic soil horizon have a major influence
on the spatial variations of the end-of-season
thaw depth;
the influence of different factors on the thawing
process is seasonally specific.
TDL = 1.412DDT0'5 + 20. I 1
R2 = 0.913 /*
/ TDD = 0.1'3D~o's + 4297
/ 0 R2= 0.043
6
.O"""
_ _ " "
"..,~,"""'"~~""
I"
/ ,
n
'I I
c
R' = 0.392
""
I
-.-I
22 27 32
DD
35 ..
0 CapeRogcahny
0 Mount Dmisy
Lavrmtla
""
""-7 j
. -.
Figure 2. End-of-season thaw depth versus square root of thawing
degree days for three CALM sites in northeast Russia.
REFERENCES
Brown, J., Hinkel K.M. & Nelson F.E. 2000. The Circumpolar Active
Layer Monitoring (CALM) program: research designs and initial
results. Polar Geography 24(3): 166-258.
Hydra soil moisture probe user's manual. Version 1.2. 1994. Vitel,
Inc., Chantilly, VA, 24 pp.
Zamolodchikov D.G., Karelin D.V., lvaschenko A.I., Oechel W.C &
Hastings S.J. in press. COz flux measurements in Russian Far
East tundra using eddy covariance and closed chamber techniques.
Tellus. Special Issue.
ACKNOWLEDMENT
The fieldwork at the Lavrentia site was supported by the
National Science Foundation (USA), Arctic Systems Science,
Land-Atmosphere-lce-Interactions Program (OPP-
9216109), as part of flux studies in Arctic ecosystems
(Zamolodchikov et al. 2003).

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Pedogenesis of carga thermochrone paleosols in North-East Russia
0. G. Zanina and S. Vi Gubin
Soil Cryology Laboratory, Insfifute of Physicochemical and Biological Problems in Soil Science, Russian Academy of Sciences,
Pushchino, Moscow Region, Russia
INTRODUCTION
Epigenetic pedogenesis
Paleosols of early Carga pedocomplex. Early Carga pa-
Late Pleistocene soils buried in ice-loess deposits were used leosols were formed during an initial thermochrone stage. The
for reconstructing environments of the Carga thermochrone. age of soils exceeds 37 thousand years. Paleosols of this
The Carga thermochrone ( 50 to 28 thousand years BP) is a period do not have the attributes of wood-soil formation. Their
period of latePleistocenetime. In theNortheastofRussia profiles are well differentiated into genetic horizons. Large
paleosols developed during this period. These paleosols are wood remains are found within humic horizons of these pacharacterised
by a complex organisation; the deposits contain leosols.
epigenetic and synlithogenetic soils (cryopedolite). Epigenetic Paleosols of late Carga pedocomplex. Soil formation took
pedogenesis in the period of the Late Pleistocene was repeat- place during later periods of Carga time, 35 to 28 thousand
edly replaced by synlithogenetic pedogenesis. The terms years ago. Pedocomplexes include three buried soils of dif-
“synlithogenetic” and “epigenetic” are used for the description ferent ages. Structures of paleosol profiles which were formed
of soil developments during this period. Epigenetic soil forma- during the initial stage of pedogenesis within the pedocomplex
tion refers to processes where differentiation of the active
(the first and the second soils) are best developed. Paleosols
layer material takes place in genetic horizons and soil profiles are represented by Histosols and Gleysols, which are didevelop.
Synlithogenetic soil formation occurs when material vided by layers of cryopedolite.
of deposits is transformed by pedogenetic processes in the The organic horizons of Gleysols differ with respect b
absence of soil profiles. The different approaches for studying structure, properties and thickness. In these soils the hydrothese
layers are determined by their genesis.
morphic characteristics and gleying are well expressed. The
thickness of profiles varies between to 1 2 m. The structure d
organic horizons in moss and sedgy peat is different. The up
METHODS
per (third) soil has badly marked structures. It was formed
under drier and even arid conditions of a harsh climate.
Buried soils contain a set of fossil organic materials in their
organic horizons (peatland, humic-peatland). Conditions d
soil formation are reconstructed using traditional paleosol
methods on the basis of analyzed organic remains and of
other soil properties. For studying synlithogenetic soils, detritus,
spores and pollen existing in the deposits are analysed.
Fossil rodent burrows of Carga cryopedolite are most infor-
mative for reconstruction due to their high content in biological
matter.
OBJECTS
Paleosols were investigated in eight outcrops of the lower
course of the Kolyma River. As a result of long-term research
in the Kolyma lowland, buried soils including the two
pedocomplexes - early Carga complex and late Carga
complex - were determined (Gubin 1994). Layers of loess
with attributes of synlithogenetic soil formation are covered by
buried epigenetic soils.
The organic horizons often consist of shallow layers
dry peat with high amounts of mineral material. The paleosol
properties of the late Carga pedocomplex show a gradual
trend towards harsher climatic conditions in this period.
Synlithogenetic pedogenesis
The cryopedolite forms under the influence of synlithogenetic
soil formation. There is no differentiation of the activelayer
material into genetic horizons. Concentrations of organic
material are low and other soil properties weakly developed.
Modern investigations of ice-loess deposits are conducted
using spore-pollen analysis along with paleontologic and paleopedologic
studies. For reconstruction of synlithogenetic soil
formation the use of these methods is not sufficient. During the
past years, ground squirrel burrows of the subgenus
Urocitellus and of other rodents (lemmings and mice) were
found in Late Pleistocene ice-loess deposits of North-East
Eurasia. These objects differ with respect to depth of location
inside the deposits, to structure and to content of frozen mate
rials. The radiocarbon age of burrows is about 32 to 28 thou-
191
d

Zanina, 0. G. et al.: Pedogenesis of carga therrnochxane paleosols ...
sand years. The formation of the burrows thus relates to the of epigenetic soil formation were characterised by a decreas-
Carga thermochrone. The analysis of contents of fossil bur- ing intensity of eolian sedimentation, a slow increase in hurows
and the ice-loess deposits containing them enables an midity and a warm climate. Synlithogenetic soil fornation
impression to be gained about the environments of corre- took place under conditions of dry tundra landscape with
sponding local habitats.
steppe vegetation as well as active accumulation of eolian
The structure of a number of fossil ground squirrel burrows deposits at the surface.
has allowed to determine the thickness of the active layer at
60 to 80 cm during the time of their construction. The presence
of sublimation ice in the central parts of chambers ACKNOWLEDGEMENTS
proves the absence of their flooding and low moisture of the
bottom parts of the active layer during maximum thaw. Uni- We thank Dr. S.V. Kiselev and Dr. G.G. Boeskorov for help
form distribution of burrows within Late Pleistocene polygons with the identification of material collected from fossil burrows.
allows to assume plairl character of a surface near the bur- This original research was supported by the Russian Founrows.
The simple structure of burrows, especially the ab dation for Basic Research, grants 01-05-64496,01-04-49087.
sene of branching and platforms of excrements close to the
entrance, points to a short-term existence of these constnrctions
and rapid blocking of the surface by deposits from active REFERENCES
eolian processes.
Fossil burrows of ground squirrel contain organic matter in Gubin, S.V., Maksimovich, S.V. and Zanina, O.G. 2001. Composition
excellent condition and obviously introduced from the surface. of seeds from fossil gofer burrows in the ice-loess deposits of
Zelyony mys as environmental indicator, Russia, Earth
Remains of higher plants, including seeds and fruits, chitin
Cryosphere, 5 (2): 76-82.
remains of insects and their larvae, eggshells, feathers, ex- Gubin S.V. 1994. The Late Pleistocene soil formation at the coastal
crements, bones and woolly mammals as well as their tis- lowlands of northern Yakutia, Russia. Soil Science 8: 5-14.
sues were found among the contents of fossil burrows. The
wool of large mammals found in burrows belongs to bisons
and musk oxen which are widely distributed in tundra-steppe
landscapes.
The amount of seeds and fruits from higher plants can reach
hundreds of thousands of units inside feeding chambers. The
great volume of fruit and seeds from fossil burrows belongs t3
the family of Brassicaceae, Ranunculaceae, Cariophy//aceae,
Poaceae and Carex. TheseedsSilene sfenophylla Ledeb,
Bisforta vivipara (L.) S.F. Gray, Potentilla nivea L and other
species of sedges were found more frequently in burrows.
The plant seeds from buried rodent burrows show the development
of predominant tundra vegetation during the Carga
thermocrone with the existence of pioneer and steppe species
(Gubin et al. 2001).
Among the remains of fossil insects, various species
such as leaf-beetle (Chrisolina rufilabris Fald, Ch. brunnicornis
bermani Medv., Galeruca dahurica Jeann), billbug
(Stephanocleonus eruditus Faust, S. fossilatus Fisch, Coniocleonus
cf. ferrugineus Fahr) and larvae of several species d
fly were discovered. These species are typical for the lower
course of the Kolyma River during Late Pleistocene times.
Almost all species are connected to steppe conditions. Collops
obscuricornis Motsch (beetle of the family of Melyridae)
was found in the west Chukotka area with modem relict
steppe. This confirms that dry steppe conditions existed the in
Northeast of Russia during Late Pleistocene time periods.
CONCLUSION
The Carga thermochrone in the investigated area is characterized
by changes from synlithogenetic to epigenetic soil
formations. Epigenetic soils buried during the Carga theme
chrone from 40 to 28 thousand years ago provide information
for reconstructing conditions of soil development. The periods
192

8th International Conference on Permafrost Extended Abstracts on Current Research and Newly Available Information
Changes in permafrost distribution and active layer th 'ckness in Canada
during the 20th century: a model assessment of clima :e change impact
Y. Zhang, W. Chen, J. Cihlar, *S.L. Smith and *D. W. Riseborough
Canada Center for Remote Sensing, Nafural Resources Canada, Offawa, Canada
*Geological Survey of Canada, Natural Resources Canada, Ottawa, Canada
INTRODUCTION
Paleoclimate records and instrumental measurements show
that the Arctic during the 20th century is the warmest in the
past four centuries, and air temperature in the northern high
latitudes has increased at a higher rate than the global mean.
The Arctic is extremely vulnerable to climatechange,and
potential impacts include changes in permafrost distribution
and active layer thickness, hydrological dynamics and carbon
sinklsource, which may positively feedback to the climate
system. This paper presents our recent results about the
transient changes in permafrost distribution and active layer
thickness in Canada during the 20th century simulated by a
process-based model at 0.5-degree latitudellongitude resolution.
scale down this monthly dataset to half-hourly intervals
(Chen et al. submitted). Other input data include land cover
types, leaf area index, surface and soil organic matter, soil
texture, the geothermal flux and ground ice content. They are
converted to 0.5-degree latitudellongitude, corresponding b
the spatial resolution of climate data. The initial conditions are
determined by iteratively running the model for an initial year
until annual soil temperature is stabilized.
METHODOLOGY
A process-based model to simulate permafrost thermal re
gimes was developed by combining the strength of existing
permafrost and land surface process models. Soil temperature
and active layer thickness (ALT) were determined by
solving the heat conduction equation, with the upper boundary
conditions being determined using the surface energy balance,
and the lower boundary conditions being defined as the
geothermal heat flux at a depth of 35 m, where heat flux is not
very sensitive to climate variations. We assumed that the
heat flux at this depth does not change with time during the Figure 1. The components of the system considered in the model,
simulation period, The model integrated the effects of climate, and energy fluxes between soil, vegetation and the atmosphere.
vegetation, soil features and hydrological conditions based on
energy and water transfer in the soil-vegetation-atmosphere
system (Fig.1). The model was validated against the meas- RESULTS AND ANALYSIS
urements at four sites in Canada. The simulation results were
in agreement with the measurements of energy fluxes, snow Figure 2 shows the simulated spatial distribution of ALT and
depth, soil temperature and thaw depth (Zhang et al. submit- its change. It also shows the distribution of permafrost areas
w .
and their change as well. The simulated permafrost distribu-
Using this model, we simulated the transient response d tion is similar to the permafrost map of Canada (Heginbottom
permafrost distribution and ALT in Canada to climate change et al. 1995), which is delineated based on permafrost site
in the 20th century. The climate data were obtained from a measurements, physiographic information and air tempera-
0.5degree latitudellongitude gridded monthly dataset from ture. (It largely represents permafrost conditions in the recent
1901 to 1995 (New et al. 2000). A scheme was developed b decades, but we show it in all the panels for spatial refer-
193

~~~~
~.~
some expansions in~iermafrost in these areas. These spatial
distribution patterns correspond to the change in patterns of air
temperature. Average ALT increased by 0.3 m for grid cells
where ALT increased, and decreased by 0.1 m for grid cells
where ALT decreased (excluding grid cells where pernabst
disappeared or formed over this period). For all the permafrost
grid cells ALT increased an average of 29% from the 1900s
to the decade 1986-1 of 995.
CONCLUSIONS
Using a process-based model and inputs from climate re
cords, soil features and remotely-sensed vegetation parameters,
the transient response of permafrost distribution and aG
tive layer thickness in Canada to climate change during the
20th century was simulated. The results show that the land
area underlain by permafrost decreased by 3% and ALT increased
by 29% during the 20th century. However, permafrost
degradation and changes in ALT differ with space and
time.
REFERENCES
Chen, W., Zhang, Y., Cihlar, J., Smith, S.L. and Riseborough, D.W.
submitted. Changes in soil temperature and active layer thickness
during the 20th century in a permafrost region in western
Canada. Journal of Geophysical Research.
Heginbottom, J.A., Dubreuil, M.A. and Harker, P.A. 1995. Canada
Permafrost. In; National Atlas of Canada 5th edition, Natural
Resources Canada, Ottawa, Canada.
New, M.G. Hulme, M. and Jones, P.D. 2000. Representing twentieth
century space-time climate variability, Part II: Development of
1901-1996 monthly terrestrial climate fields. Journal of Climate
13: 2217-2238.
Zhang, Y., Chen, W. and Cihlar, J. submitted. A process-based
model for quantifying the impact of climate change on permafrost
thermal regimes. Journal of Geophysical Research.
194

Author Index
Abramov A. A., 1, 1 15
Ahonen L., 141
Akerman J. H., I 9
Allard M., 37
Ammann W., 95
Arenson L. U., 71
Ariuntsetseg L., 45
Asano K., 13
Auerswald K., 135
Avian M., 3
Balobajev V., 5
Baron L., 89
Baroni C., 145
Barry R. G., 21, 123
Batbold A., 163
Beer C., 7
Beylich A. A,, 9
Blomqvist R., 141
Borzotta E., I I
Brabham P., 139
Brouchkov A., 13
Brown J., 169, 153, 113, 123
Buldovich S. N., 1
Burgess M. M., 153
Bykhovets S., 21
Cai M. F., 97
Campbell S., 139
Carton A,, 145
Celi L., 39
Chekhina I., 15
Chelli A., 17
Chen W., 193
Chestnykh O., 103
Christiansen H. H., I 9
Chudinova S. M., 21
Cihlar J., 193
Dankers R., 85
Davydov S. P., 41
Degnan P., 141
Delaloye R., 23, 93, 107, 89, 173
Domnikova E. A., 25
Drozdov D. S., 27,149
Ednie M., 29
Eiken T., 181
Elliott C., 31
Enkhtuul M., 163
Etzelmuller B., 45
Fabre D., 33
Federici P. R., 33
Fishback L-A., 35
Fortier D., 37
Frappe S. K., 141
Frauenfelder R., 155,45
Freppaz M., 39
Frolov A. D., 67
Fukuda M., 53,13
41,75
Gakhova E. N., 187
Gilichinsky D. A., 21,75, 115, I, 41, 147
Fyodorov-Davydov D. G.,
Gintz D., 9
Gorshkov S. P., 43
Goulden C., 45
Gruber S., 47,83, 77,65,59
Gubin S. V., 75,41, 147, 191, 187
Gubler H., 61
Gude M., 7,111
Haeberli W., 49,95, 125, 173
Hallet B., 151
Hamano Y., 157
Hanson S., 51
Harada K., 53
Harris C., 139
Hauck C., 55,57
Heggem E. S. F., 45
Heiner S., 59
Herz T., 61, 129,77,65
Hinkel K. M., 113
Hjort Jan, 63
Hoelzle M., 47, 11 1, 51, 173
Hof R., 65, 77
Hollister R. D., 165
Huggel C., 49
lshikawa M., 67
lstratov V. A,, 69
Jambaljav Ya., 45, 163
Jensen M., 141
Johansen M. M., 71
Kaab A., 155,45,183,125,49
Kadota T., 67
Karelin D., 189
Kasymskaya M., 73
Kholodov A. L., 75, 1,41

King L., 77, 129, 61, 65
King R. H., 35
Kizya kov A. I., 79
Kneisel C., 81, 55
Kobabe S., 175
Kohl T., 83
Kolstrup E., 9
Korostelev Y. V., 27
Koster E. A., 85
Kotov A,, 189
Kristensen L., 87
Krummenacher B., 173
Kuchmin A. O., 67
Kudryavtsev S., 167
Lambiel C., 89, 93
Laurinavichius K. S., 1
Lehto K., 141
Leibman M. O., 105
Lewkowicz A. G., 29
Linde N., 9
Lisyuk M., 167
Little J., 91
Lucht W., 7
Lugon R., 93,107
Luetscher M., 143
Lutschg M., 95
Ma Q., 97
Ma W., 97
Makarov V. N., 99
Malkova G., 103
I01
Marchenko S. S.,
Mazhitova G., 103
Meier E., 59
Melnikov E. S., 105
Metrailler S., 107
Mergelov N. S., 41
Mitusov A. V., I09
Mitusova 0. E., 109
Molau U., 9
Monbaron M., 23,143
Monnet R., 89
Moren L., 141
Moskalenko N. G.,
105,169,177
Mustafa O., 111
Nelson F. E., 113,91
Nikiforov S., 127
Novototskaya-Vlasova K., 115
Odegird R. S., 181
Oelke C., 117
Ogorodov S., 119
Ohata T., 67
Osokin N., 49
Ostroumov V. E., 41
Paananen M., 141
Panova Y., 121
Pappalardo M., 33, 17
Paramonov V., 167
Parsons M. A., 123
Paul F., 125
Pavlidis Y., 127
Pedersen L. B., 9
Perednya D. D., 79
Perekalin S. O., 67
Pfeiffer E."., 175
Philippi S., 129
Phillips M., 39, 135
Polkvoj A,, 49
Popova Alexandra A,, 131
Puigdomenech I., 141
Rachold V., 127
Raynolds M. K., 177
Razzhivin V., 189
Repelewska-Pekalowa J., 19
Reynard E., 93
Ribolini A., 133, 33, 17
Rist A. , 135
Riseborough D., 193
Rivkin F. M., 15
Rivkina E. M., 75, I
Rodriguez J. A. P., 137
Romanovskii N., 73
Romanovsky V., 153,45
Ross N., 139
Ruskeeniemi T., 141
Sandall H., 91
Saruul N., 45
Sasaki S., 137
Seppala M., 63
Seppi R., 145
Sergueev D., 73
Serrano E., 93
Schlapfer D., 47
Schlatter F., 143
Schmullius C., 7
Schnellman M., 141
Sharkuu N., 67,45
Shashkin K., 167
Shatilovich A., 147
Shcherbakova V. A., 1
Shepelev V. V., 5
Shiklomanov N. I., 113
Shmakova L. A., 147
149
Sletten R. S., 151
Skvortsov A. G.,

Smith S. L., 153, 193
Soina V., I 15
Sollid J. L., 181
Sorokovikov V. A., 21 , 7541
Stockli V., 39
Stoeckli V., 95
Stoffel M., 143
Streletskaya I. D., 159
Strozzi T., 155
Sueyoshi T., 157
Tanaka M., 13
Tarasov A., 159
Thyrsted T., 9
Tishkova M., 43
Tomita F., 13
Trombotto D. T., 161, 11
Tumentsetseg S., 45
Tumurbaatar D., 163
Tweedie C. E., 165
Ukraintseva N. G., 159
Ulitsky V., 167
Undarmaa D., 163
van der Linden S., 85
Vasiliev A. , 169, 105
Venevsky S., 171
Vonder Muhll D., 173
Wada K., 53
Wagner D., 175
Wagner U., 57
Walegur M., 91
Walker D. A., 177
Walker H. J., 179
Wangensteen B., 181
Warren I. M. T., 185
Webber P. J., 165
Weber M., 183
Wegmiiller U., 155
Wikstrom L., 141
Williams P. J., 185
Yashina S, G., 187
Zamolodchikov D., 189, 103
Zanina 0. G., 191
Zhang T., 21 , 117,123
Zhang Y., 193,67
Zhang Z. H., 97
Zimov S. A., 41
Zotikov I., 49